Ieee Icma 2020

Home , Chen Jiao, Daisuke Sasaki, Guo Jian, Huang Qiang (politician), Jingxiang, Liu Yijun (guitarist)

Conference Digest

2020 IEEE International Conference on Mechatronics and Automation

IEEE ICMA 2020

Beijing, China

October 13-16, 2020

Cosponsored by

IEEE Robotics and Automation Society Beijing Institute of Technology, China Kagawa University, Japan

Technically cosponsored by

National Natural Science Foundation of China, China Chinese Mechanical Engineering Society Chinese Association of Automation State Key Laboratory of Robotics and System (HIT) The Institute of Advanced Biomedical Engineering System, BIT Key Laboratory of Convergence Medical Engineering System and Healthcare Technology, the Ministry of Industry and Information Technology, Beijing Institute of Technology, China Guangxi University of Science and Technology Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems; Tianjin Key Laboratory for Advanced Mechatronics System Design and Intelligent Control; Tianjin University of Technology, China Tianjin International Joint Research and Development Center, Tianjin University of Technology, China Harbin Engineering University The Robotics Society of Japan The Japan Society of Mechanical Engineers Japan Society for Precision Engineering The Society of Instrument and Control Engineers University of Electro-Communications University of Electronic Science and Technology of China Optics and Precision Engineering

IEEE ICMA 2020 PROCEEDINGS

Additional copies may be ordered from:

IEEE Service Center 445 Hoes Lane Piscataway, NJ 08854 U.S.A.

IEEE Catalog Number: CFP20839-PRT ISBN 978-1-7281-6415-1 IEEE Catalog Number (CD-ROM): CFP20839-CDR ISBN (CD-ROM): 978-1-7281-6414-4

Copyright and Reprint Permission: Abstracting is permitted with credit to the source. Libraries are permitted to photocopy beyond the limit of U.S. copyright law for private use of patrons those articles in this volume that carry a code at the bottom of the first page, provided the per-copy fee indicated in the code is paid through Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923. For reprint or republication permission, email to IEEE Copyrights Manager at [email protected]. All rights reserved. Copyright ©2020 by IEEE.

The papers in this book comprise the proceedings of the meeting mentioned on the cover and title page. They reflect the author’s opinions and, in the interest of timely dissemination, are published as presented and without change. Their inclusion in this publication does not necessarily constitute endorsement by the editors or the Institute of Electrical and Electronics Engineers, Inc.

The Institute of Electrical and Electronics Engineers, Inc.

Foreword

On behalf of the IEEE ICMA 2020 Conference Organizing Committee, it is our great pleasure, an honor, and a privilege to welcome you to Beijing for the 2020 IEEE International Conference on Mechatronics and Automation. This conference reflects the growing interests in the broad research areas of mechatronics, robotics, sensors and automation.

ICMA 2020 marks the 17th edition of the IEEE ICMA annual conference series. We are proud to announce that a high number of 529 papers were submitted from 28 countries and regions, including 511 contributed papers, 18 papers for organized sessions, and 342 papers were accepted for oral or poster presentation at the conference after a rigorous full-paper review process, achieving an acceptance rate of less than 65%. Presentations at ICMA 2020 are organized in 6 parallel tracks, for a total of 55 sessions, including 1 poster session, taking place during the three conference days. We are fortunate to be able to invite four distinguished speakers to deliver plenary talks.

We are very pleased that you are joining us at IEEE ICMA 2020 in Beijing to take part in this unique experience. The main objective of IEEE ICMA 2020 is to provide a forum for researchers, educators, engineers, and government officials involved in the general areas of mechatronics, robotics, sensors and automation to disseminate their latest research results and exchange views on the future research directions of the related fields. IEEE ICMA 2020 promises to be a great experience for participants from all over the world, with an excellent technical program as well as social activities.

We would like to express our most sincere appreciation and thanks to all of our sponsoring societies and organizations and to all the individuals who have contributed to the organization of this conference. Our special thanks are extended to our colleagues in the Program Committee for their thorough review of all the submitted papers, which is vital to the success of this conference. We must also extend our thanks to our Organizing Committee and our volunteers who have dedicated their time toward ensuring the success of this conference. Last but not least, we thank all the contributors for their support and participation in making this conference a great success. Finally, we wish you a great conference and enjoyable stay in Beijing, China.

Qin Li Hideyuki Sawada Liwei Shi Shuxiang Guo Xueshan Gao General Chair General Chair Program Chair Organizing Chair Organizing Chair

Welcome Remarks

It is my pleasure to welcome you to join us at the 2020 IEEE International Conference on Mechatronics and Automation (IEEE ICMA 2020) on behalf of Beijing Institute of Technology. We are delighted to host the 17th edition of the IEEE ICMA Conference of the annual series, which reflect the growing interests in the broad research areas of mechatronics, robotics, sensors and automation. To begin with, I would like to make a brief introduction to Beijing Institute of Technology. Beijing Institute of Technology (BIT) is one of the national key universities in China as an open, public, research-oriented university with a focus on science and technology. As the 10th university to enter the “985 Project” and as one of the first 15 universities to be acknowledged into “211 Project” which aim at establishing world renowned universities, BIT’s annual research fund in 2019 has amounted to over 2 billion RMB ranking top 10 in China. And the research on mechatronics and automation at BIT has become increasingly active. It is sincerely hoped that IEEE ICMA 2020 will provide a forum for researchers, educators, engineers, and government officials involved in the broad research areas of mechatronics, robotics, sensors and automation to exchange ideas on their latest research and views on the future research directions of the related fields. Finally, I would like to express my sincere gratitude to all of the sponsoring societies and organizations as well as all the individuals contributed to organizing the Conference. Also, special thanks go to all the authors, session organizers, plenary speakers, exhibitors for their research works and making IEEE ICMA 2020 a successful and fruitful event. To all participants, I extend my heartfelt welcome and thanks for attending this event.

Jun Zhang, Professor President, Beijing Institute of Technology Welcome Message

It is my honor to welcome you to attend the 2020 IEEE International Conference on Mechatronics and Automation (IEEE ICMA 2020) on behalf of Guangxi University of Science and Technology. We are delighted to support the Conference which marked as the 17th edition of the IEEE ICMA among the annual conference series. The Conference reflects the growing interests in the broad research areas of mechatronics, robotics, sensors and automation. I would like to make a brief introduction to Guangxi University of Science and Technology. Guangxi University of Science and Technology is located in Liuzhou City, the cradle of Ancient Humans called Liujiang man in southern China, the pillar of Guangxi Zhuang Autonomous Region, as well as the second largest city in Guangxi Zhuang Autonomous Region. Today, Liuzhou has been a noted city with ecological and livable civic environment, excellent tourism, and rich historical and cultural resources. It is sincerely hoped that IEEE ICMA 2020 will provide a forum for researchers, educators, engineers, and government officials involved in the general areas of mechatronics, robotics, sensors and automation to disseminate their latest research results and exchange views on the future research directions of the related fields. Finally, I would like to express my sincere gratitude to all of the sponsoring societies and organizations as well as all the individuals contributed to the organization of the Conference. Also, special thanks are owed to all the authors, session organizers, plenary speakers, exhibitors for contributing their research works and making IEEE ICMA 2020 a successful and fruitful event. To all participants, I extend my heartfelt welcome and thanks for attending this event, wish your stay here in Beijing, China, is very pleasant and enjoyable.

Simin Li, Professor President Guangxi University of Science and Technology

IEEE ICMA 2020 Conference Digest Table of Content

Foreword Welcomg Remarks

IEEE ICMA 2020 Conference ...... i Organizing Committees ...... i International Program Committee ...... iv IEEE ICMA 2020 Conference Cosponsors ...... xi General Information ...... xii Conference Information ...... xv Conference Registration ...... xv Social Events ...... xv Technical Program Plenary Talks ...... xvi Program at a Glance ...... xxv Technical Session Schedule ...... xxvi Technical Sessions Wednesday, 14 Obtober 2020 ...... 1 Thursday, 15 Obtober 2020 ...... 27 Friday, 16 Obtober 2020 ...... 51 Index of Session chairs ...... 61 Index of Authors ...... 64 IEEE ICMA 2021 CFP & IJMA Journal CFP

IEEE ICMA 2020 Conference

International Scientific Advisory Board

Advisory Council Honorary Chairs T. J. Tarn Washington University, USA Toshio Fukuda Meijo University, Japan Advisory Council Chairs Tianyou Chai Northeastern University, China Hegao Cai Harbin Institute of Technology, China Liding Wang Dalian University of Tech., China A.A. Goldenberg University of Toronto, Canada Kazuhiro Kosuge Tohoku University, Japan Paolo Dario Scuola Superiore Sant'Anna, Italy Masayoshi Tomizuka UC Berkeley, USA Mario A. Rotea University of Massachusetts, USA Ju-Jang Lee KAIST, Korea Ren C. Luo National Taiwan University, Taiwan Yu Yao Harbin Engineering University, China Yanrong Li UESTC, China Huadong Yu CUST, China Qingxin Yang Tianjin University of Technology, China

Organizing Committees

General Chairs

Hideyuki Sawada Waseda University, Japan Qin Li Beijing Institute of Technology, China

i General Co-Chairs

Shigeki Sugano Waseda University, Japan William R. Hamel University of Tennessee, USA Qiang Huang Beijing Institute of Technology, China Sergej Fatikow University of Oldenburg, Germany Darwin G. Caldwell Italian Institute of Technology, Italy Yuxin Zhao Harbin Engineering University, China Lixin Dong City University of Hong Kong, China Zhan Yang Suzhou University, China Simin Li Guangxi Univ. of Science and Tech., China Baofeng Zhang Tianjin University of Technology, China Yuanqing Xia Beijing Institute of Technology, China

Organizing Committee Chairs

Shuxiang Guo Kagawa University, Japan Jian Sun Beijing Institute of Technology, China Xueshan Gao Beijing Institute of Technology, China

Organizing Committee Co-Chairs

Aiguo Ming Univ. of Electro-Communications, Japan Qinxue Pan Beijing Institute of Technology, China Hideyuki Hirata Kagawa University, Japan

Tutorial/Workshop Chairs

Guangjun Liu Ryerson University, Canada Huosheng Hu University of Essex, U.K

Invited/Organized Session Chairs

Hiroyuki Nabae Tokyo Institute of Technology, Japan Shingo Maeda Shibaura Institue of Technology, Japan Meisuke Morishima Osaka University, Japan

ii Chaoyang Shi Tianjin University, China Yongfu Li City University of Hong Kong, China Yasuhisa Hirata Tohoku University, Japan

Awards Committee Co-Chairs

James K. Mills University of Toronto, Canada Qiang Fu Tianjin University of Technology, China

Publications Chairs

Jin Guo Beijing Institue of Technology, China

Publicity Chairs

Jinjun Shan York University, Canada Xiufen Ye Harbin Engineering University, China

Finance Chair

Jian Guo Tianjin University of Technology, China Hidenori Ishihara Kagawa University, Japan

Local Arrangement Chairs

Yi Xin Beijing Institute of Technology, China Xiao Zhang Beijing Institute of Technology, China Yan Zhao Hebei University of Technology, China

Conference Secretariats

Y. Liu, Z. Wang, R. An, X. Jin Kagawa University, Japan P. Shi, Z. Yang, L. Zheng Kagawa University, Japan C. Yang, W. Zhou, D. Bu Beijing Institute of Technology, China Xiaolong Yang Guangxi Univ. of Science and Tech., China

Conference Web System Administrator

Yi Liu Kagawa University, Japan

iii Program Committees

Program Chair

Liwei Shi Beijing Institute of Technology, China

Program Co-Chairs

Kohei Nakajima The University of Tokyo, Japan Ryuma Niiyama The University of Tokyo, Japan Tatsuo Arai Osaka University, Japan Kevin Lynch Northwestern University, USA Cecilia Laschi Scuola Superiore Sant'Anna, Italy Stefan Byttner Halmstad University, Sweden Wan Kyun Chung POSTECH, Korea Yili Fu Harbin Institute of Technology, China Jian Li Guangxi Univ. of Science and Tech., China Anqi Qiu National Univer. of Singapore, Singapore Yajing Shen City University of Hong Kong, China Zhangguo Yu Beijing Institute of Technology

International Program Committee

Adachi, Fumiyuki Addie, Ron Aiyama, Yasumichi Althoefer, Kaspar An, Ruochen Ando, Kazuaki Ang, Wei Tech Aoyama, Hisayuki Arai, Fumihito Arai, Tamio Arai, Tatsuo Araki, Kiyomichi Arzanpour, Siamak Asada, Minoru Asama, Hajime Asano, Toshio Baeg, Sang Hyeon Bai, Ou Bao, Xianqiang Ben, Ridha Ben-Tzvi, Pinhas Bi, Shusheng Bi, Zhuming Bian, Hongyu Bidaud, Philippe Bolon, Ph. Bolon, Philippe Boustany, Nada Brown, Martin Bu, Dongdong Byung, Min Cai, Lilong Caldwell, Darwin G. Cao, Maoyong Ceccarelli, Marco Chen, Chun-Ta Chen, Deyun Chen, Feng Chen, Guanlong Chen, I-Ming

iv Chen, Ken Chen, Liguo Chen, Ting Chen, Weidong Chen, Weihai Chen, Wenhua Chen, Xin Chen, Xinkai Chen, Yangquan Chen, Zhangwei Cheng, Allen Cheng, Jianhua Cheng, Ji-Xin Cheol, Min Cheong, Joono Cho, Young-Jo Choi, Hyouk Ryeol Choi, Hyun-Taek Choi, Junho Choi, Youngjin Chou, Wusheng Chu, Jinkui Chugo, Daisuke Chui, Dehua Chung, Woojin Cui, Jianwen Dai, Jiansheng Dai, Xuefeng Dailey, Matthew Damaren, Chris Dario, Paolo Davis, Steve Dechev, Nick Deco, Gustavo Deguang, Shang Demiris, Yiannis Deng, Yulin Dillmann, Ruediger Ding, Jiexiong Ding, Xilun Do, Hyun Min Dodd, Tony Doh, Nakju Dohta, Shujiro Doi, Shunichi Dong, Enzeng Dong, Hongbin Dong, Zaili Du, Mingxing Du, Pingan Du, Xiliang Du, Zhijiang Duan, Haibin Dubay, Rickey Dufosse, Michel Fan, Jinwei Fang, Hao Fang, Yongchun Fei, Qin Feng, Gary Feng, Weixing Fiorini, Paolo Foulloy, Laurent Fu, Bin Fu, Jiacai Fu, Mingyu Fu, Qiang Fu, Sheng Fu, Shuigen Fu, Yili FUH, Ying-His Fujino, Tadashi Fujisaki, Kiyotaka Fujisawa, Shoichiro Fujiwara, Takayuki Fujiwara, Yoshitaka Fukuda, Toshio Fukuyama, Hidenao Fung, Wai-Keung Furusho, Junji Furuta, Katsuhisa Furutani, Eiko Furutani, Katsushi Furuya, Nobuyuki Gang, Tong Gao, baofeng Gao, Guohua Gao, Hongtao Gao, Lin Gao, Shesheng Gao, Wei Gao, Xueshan Ge, Sam Ge, Sam Shuzhi Ge, Weimin Ge, Yunjian Gitchang Gofuku, Akio Gong, Haixia Gong, Yadong Gong, Zhiguo Graefe, Volker Gu, Dongbing Gu, Jason Gu, Xingzhong Guan, Yisheng Guang, Zu Guang Guglielmelli, Eugenio Guo, Jian Guo, Jin Guo, Maozu Guo, Mingliang Guo, Peng Guo, Shuxiang Guo, Yi Habibi, Saeid Haga, Yoichi Hamaguchi, Tetsuya Hamel, William R. Han, Hongbin Han, Jiqing Han, Min Han, Yujie Hane, Kazuhiro Hao, Gang Harada, Kensuke Hariri, Alireza Hasegawa, Osamu v Hasegawa, Tadahiro Hasegawa, Yasuhisa Hashimoto, Hiroshi Hashimoto, Koichi Hashimoto, Minoru Hashimoto, Shuji Hata, Seiji Hattori, Tetsuo Hayashi, Jun-ichiro He, Cunfu He, Jia He, Jiping Hino, Junichi Hirai, Shigeoki Hirata, Hideyuki Hirata, Koichi Hirata, Yasuhisa Hirose, Akira Hong, Keum-Shik Hong, Seung Ho Hori, Toshio Hosoda, Kou Hou, Mingshan Hou, Shuping Hou, Xihuan Hou, Zhenguang Hu, Chao Hu, Huosheng Hu, Jiquan Hu, Jun Hu, Jwu-Sheng Hua, Jun Huang, Dagui Huang, Guo-Shing Huang, Hongzhong Huang, Ping Huang, Qiang Huang, Qingjiu Huang, Tian Huang, Yalou Huasong, Min Hwan, Chang Hwan, Dong IBA, Hitoshi Ichikawa, Akihiko Ichiki, Masatoshi Iguchi, Tetsuhiro Ikehara, Masaaki Ikeuchi, Masashi Ikuta, Koji Inomo, Hitoshi Ise, Toshifumi ishiguro, Hiroshi Ishihara, Hidenori Ishihara, Sunao Ishii, Akira Ishii, Chiharu Ishii, Kazuo Ishii, Koji Ishikawa, Junzo Ito, Tomotaka Iwamura, Yoshiaki Iwatsuki, Nobuyuki Izuishi, Kunihiko Janabi-Shari, Farrokh Ji, Ping Ji, Yuehui Jia, Kebin Jia, Songmin Jia, Zhengyuan Jiang, Pingyu Jiang, Zhen Jiao, Shuhong Jin, Hongzhang Jin, Rencheng Jing, Wuxing Jo, Yongho Jr., Marcelo H.Ang, Jung, Kwangmok Jung, Seul K.H.Pang, Grantham Kagami, Shingo Kagawa, Koji Kamata, Minoru Kamiya, Yoshitsugu Kanamori, Chisato Kaneko, Shunichi Kang, Sung Chul Kang, Taehun Karaki, Masayuki Kato, Takahisa Khajepour, Amir Kiguchi, Kazuo Kim, Byeongho Kim, Doik Kim, Jinhyun Kim, Jung Kim, Keehoon Kim, Sungshin Kim, Wheekuk Kimura, Kotaro Kitajima, Hiroyuki Kitajima, Katsuhiro Kobayashi, Tetsuo Kobayashi, Toshihiro Koh, Kyungchul Kojima, Masaru Kok, Tan Komeda, Takashi Kometani, Reo Kong, Xiawen Konyo, Masashi Koo, Ja Choon Koshimizu, Hiroyasu Kosuge, Kazuhiro Kotiw, Mike Kotosaka, Shinya Kouzani, Abbas Koyanagi, Kenichi Koyanagi, Mitsumasa Kubota, Naoyuki Kubota, Takashi Kulkarni，M. S. Kuno, Yoshinori Kurazuma, Ryo Kurisu, Masamitsu Kuroki, Nobutaka Kuwakado, Hidenori vi

Kuwano, Hiroki Kyriakopou, Kostas J. Lai, Yongjun Lan, Hai Laschi, Cecilia Laugier, Christian Lee, Jang Myung Lee, Jihong Lee, Songjun Lee, Yikuen Li, Bin Li, Bing Li, Chunwen Li, Desheng Li, Gang Li, Haisen Li, Hongsheng Li, Howard Li, Jianfeng Li, Jiangang Li, Jin Li, Li Li, Mantian Li, Maoxun Li, Qin Li, Wenfeng Li, Xiaoshan Li, Xiaoyang Li, Xiukun Li, Yangmin Li, Youfu Li, Yuhua Li, Yun Li, Yunhua Li, Zhang Li, Zhiyi Li, Zhiyong Li, Zhongjian Lian, Feng-Li Liang, Guolong Liang, Yan Liao, Hongen Liao, Wei-Hsin Lin, Chyi-Yeu Lin, Hsiung-Cheng Lin, Ming-Tzong Liu, Bo Liu, Da Liu, Dikai Liu, Fang Liu, Fanming Liu, Guangjun Liu, Guangyu Liu, Haibo Liu, Honghai Liu, Hugh Liu, Jiming Liu, Jindong Liu, Jinguo Liu, Jun Liu, Liqiang Liu, Ming Liu, Peter X. Liu, Qing Liu, Rong Liu, Xiangdong Liu, Xiaoping Liu, Xin-Jun Liu, Yongguang Liu, Yugang Liu, Yunhui Liu, Zhen Long, Guilu Lottin, Jacques Lou, Yunjiang Loureiro, Rui Low, Kin-Huat Lu, BaoLiang Lu, Shengfu Lu, Tien-Fu M.Chen, Ben M.Gupta, Madan Ma, Chunguang Ma, Shugen Ma, Xu Ma, Xudong Mae, Yasushi Magnenat-Thalm, Nadia Mao, Jianqin Maruyama, Hisataka Masaki, Yamakita Masek, Vlastimil Mashatan, Vahid Mashec, Vlasitimi Masuda, Tadashi Matsuhisa, Hiroshi Matsunaga, Saburo Matsuno, Fumitoshi Matsuno, Takayuki Matsushige, Kazumi Maxwell, Andrew Melchiorri, Claudio Meng, Max Q.-H. Meng, Yan Mills, James K. Mills, James K. Minami, Hirotsugu Minato, Kotaro Minemura, Kiyoshi Ming, Aiguo Mitsuishi, Mamoru Miyanaga, Yoshikazu Miyauchi, Satoru Mo, Hongwei Mo, Shuhua Mochiyama, Hiromi Morii, Masakatsu Morikawa, Hiroyuki Morishige, Koichi Morishima, Keisuke Morita, Noboru Morita, Yoshifumi Murakami, Toshiyuki Muscato, Giovanni Nagata, Fusaomi Nagatani, Keiji Nagato, Keisuke Nakajima, Masahiro vii Nakamura, Akio Nakamura, Hikaru Nakao, Masayuki Nakatani, Akihiro Nakauchi, Yasushi Nanayakkara, Thrish Nefti-Meziani, Samia Nelson, Bradley J. Ni, Jinping Ni, Zhonghua Ning, Hui Nohmi, Masahiro Nojima, Toshio Oana, Hidehiro Obara, Minoru Ogose, Shigeaki Ohara, Kenichi Ohsawa, Yasuharu Ohtake, Hiroshi Oka, Koichi Okada, Eiji Oki, Eiji Okuma, Masaaki Omichi, Takeo Oohira, Fumikazu Osumi, Hisashi Otake, Mihoko Otsuka, Akimasa Ouezdou, Fathi Ben Ouyang, Puren P.Miller, David Pan, Yajun Pang, Muye Park, Jong Hyeon Park, Jooyoung Park, Sangdok Payande, Sharam Perez, Ruben Pobil, Angel P. del Prassler, Erwin Qi, Guangyun Qi, Hairong Qi, Naiming Qiao, Gang Qiao, Hong Qiu, Anqi Qiu, Hua Radermacher, Klaus Rao, Wenbi Ren, Carolyn Ren, Wei Ren, Xiangshi Rhim, Sungsoo Ridha Ben Mrad Roh, Segon Rong, Weibin Ru, Changhai Ryeol, Dong Ryu, Jee-Hwan Sabatier, Jocelyn Sabti, Ali Saito, Takashi Sakaguchi, Masamichi Sakai, Shuichi Sakka, Sophie Sakurai, Yoshio Salman, Shaaban Ali Sampei, Seiichi Sandini, Giulio Sawada, Hideyuki Sekiyama, Kousuke Semini, Claudio Shaaban Ali Salman Shan, Jinjun Shao, Keyong Shao, Xinjian Shen, Yantao Shen, Yueshi Sheng, Jie Sheng, Weihua Shi, Chaoyang Shi, Guangfan Shi, Haizhang Shi, Jichuan Shi, Liwei Shi, Zhen Shibata, Takanori Shimojo, Makoto Shimotsu, Masateru Shoham, Moshe Shriaki, Wataru Singh, Akash Soar, Jeffrey Son, Jaebum Song, Jae-Bok Song, Kai-Tai Song, Quanjun Song, Yu Song, Zhibin Stasse, Olivier Stein, Cathryne Su, Chanmin Q. Su, Chanmin Q. Su, Chun-yi Su, Liying Sugar, Tom Sugita, Naohiko Suh, Il Hong Sun, Baoyuan Sun, Daqian Sun, Dong Sun, jinwei Sun, Kangning Sun, Xiaojun Sun, Yong Sun, Yu Sun, Zhaowei Suzuki, Keisuke Suzuki, Minoru Suzuki, Takahiro Suzuki, Yuji Tadakuma, Kenjiro Takahashi, Ryoichi Takahashi, Satoru Takahashi, Satoshi Takahashi, Tatsuro Takaiwa, Masahiro Takamasu, Kiyoshi viii

Takanobu, Hideaki Takasaki, Masaya Takeda, Takashi Takeda, Yukio Takesue, Naoyuki Takubo, Tomohito Tan, Jeffrey Tan, Jindong Tan, Lihai Tan, Min Tan, Zhenfan Tanaka, Mami Tanaka, Takayuki Tang, Mo Tang, Yike Tanikawa, Tamio Tanji, Yuichi Tao, Nongjian Tarumi, Hiroyuki Terada, Hidetsugu Tian, Yantao Tomita, Naohide Tonet, Oliver Torii, Akihio Touge, Tetsuo Tsagarakis, Nikos Tsai, Ching-Chih Tsuji, Toshio Tsukada, Toshihiko Tsukamoto, Hiroshi Tsumaki, Yuichi Tsunoda, Okitoshi Tung, Steve Ueno, Satoshi Vachkov, Gancho Vai, Ming-I Vanderborght, Bram Vernon, Brent Vlacic, Ljubo Voos, Holger Wada, Masayoshi Wada, Osami Wada, Takahiro Wan, Feng Wan, Xinhua Wang, Baikun Wang, Cheng Wang, DongMei Wang, Gang Wang, Guoli Wang, Hua Wang, Joseph Wang, Keqi Wang, Lizhen Wang, Ludan Wang, Michael Yu Wang, Peishan Wang, Pu Wang, Shuchen Wang, Shuxing Wang, Tongyu Wang, Wen Wang, Xianghong Wang, Xiaoyun Wang, Xin Wang, Xinsong Wang, Yafeng Wang, Yuechao Wang, Yuezong Wang, Zhidong Wang, Zhuo Wang, Zongyi Wang, Zuobin Warisawa, Shin-ichi Watanabe, Keigo Watanabe, Mutsumi Wei, Wei Wen, Bangchun Wen, Paul Wong, Pak-Kin Wu, Changhua Wu, Gang Wu, Jinglong Wu, Lei Wu, Qiong Wu, Shijing Wu, Xiaofeng Wu, Xiaojun X.Yang, Simon Xi, Jeff Xi, Zhihong Xiang, Zhengrong Xiao, Jizhong Xiao, Lan Xiao, Nan Xie, Lihua Xie, Ming Xie, Shane Xie, Shaorong Xin, Ming Xing, Huiming Xiong, Caihua Xu, Bo Xu, Chunquan Xu, De Xu, Dingjie Xu, Fen Xu, Honghai Xu, Jianan Xu, Lixin Xu, Mengguo Xu, Qingsong Xu, Shijie Xu, Yaoqun Xue, AnKe Xue, Dingyu Yakou, Takao Yamada, Takayoshi Yamaguchi, Tomomi Yamamoto, Manabu Yamamoto, Motoji Yamamoto, Yoshio Yamashita, Atsushi Yamaura, Hiroshi Yan, Shaoze Yan, Shengyuan Yan, Zhao Yanagihara, Mamoru ix

Yang, Cheng Yang, Enxia Yang, Erfu Yang, Fang Yang, Guiliin Yang, Hyun Suck Yang, Jianwu Yang, Jing Yang, Kwangjin Yang, Qingsheng Yang, Wu Yang, Xiukun Yang, Yong Yang, Yousheng Yang, Zhaojun Yano, Masafumi Yao, Yiyu Ye, Cang Ye, Changlong Ye, Shujiang Ye, Xiufen Yi, Byung-Ju Yi, Chuanyun Yi, Jianqiang Yin, Guofu Yin, Zhengsheng Yin, Zhouping Ying, Lixia Ying, Xianghua Yokoi, Kazuhito Yokokohji, Yasuyoshi Yokota, Sho Yoshida, Shunichi You, Bo Young, Nak Yu, Dejie Yu, Huadong Yu, Jie Yu, Junzhi Yu, Qiang Yu, Shui Yu, Xiaoyang Yu, Yong Yu, Yueqing Yu, Zhangguo Yuan, Jianjun Yuan, Juntang Yuan, Libo Yuan, Xiaobu Yue, Chunfeng Yue, Dong Yue, Yong Yun, Chao Yuta, Shinichu Zeng, Chunnian Zha, Hongbin Zhang, Baida Zhang, Chengjin Zhang, Dan Zhang, Dianlun Zhang, Hong Zhang, Jianpei Zhang, Jianwei Zhang, Jinxiu Zhang, Lei Zhang, Lijun Zhang, Lixun Zhang, Mingjun Zhang, Rubo Zhang, Sen Zhang, Songyuan Zhang, Xianmin Zhang, Xiaolong Zhang, Xiaoyu Zhang, Xinming Zhang, Xuping Zhang, Yanhua Zhang, Yi Zhang, Yimin Zhang, Yong Zhang, Yongde Zhang, Yonggang Zhang, Youmin Zhang, Yunong Zhang, Zhaohui Zhang, Zhe Zhao, Cangwen Zhao, Chunhui Zhao, Lin Zhao, Qing Zhao, Xin Zhao, Xinhua Zhao, Yuxin Zhao, Zhijun Zheng, Fei Zheng, Guibin Zheng, Jinyang Zheng, Liang Zheng, Lingling Zheng, Yuanfang Zhong, Ning Zhou, Wei Zhou, Xunyu Zhu, Chi Zhu, Chunbo Zhu, George Zhu, Jianguo Zhu, Qidan Zhu, Xiangyang Zhu, Xiaorui Zhu, Xilin Zhu, Yu Zu, Jean Zyada, Zakarya

x IEEE ICMA 2020 Conference Cosponsors

Cosponsored by

IEEE Robotics and Automation Society Beijing Institute of Technology, China Kagawa University, Japan

Technically cosponsored by

General Information

Beijing

Beijing, which was founded 3000 years ago, is the capital of the People's Republic of China (PRC). It is also the nation's political and cultural hub. Additionally, it is the focal point for the country's transportation, scientific and technological development, education and communication. Its present-day population is over eleven million, thus, it is the second largest city in China. Previously known in English as Peking, the name was changed when the system for spelling Chinese words in English changed; the name in English means "northern capital". Beijing is one of the Great Ancient Capitals of China and has hosted the seat of government for much of China's history. It is the political, economic, academic, and cultural center of the country. Tradition and modern civilization are well integrated in this beautiful city.

The long history of Beijing endows the city with a rich cultural heritage. The Great Wall, one of the world's great wonders and one of the very few man-made structures that can be seen from space, extends several thousand miles, and passes relatively near to Beijing. The Forbidden City includes the most splendid group of imperial palaces in the world. The temple of heaven is the place of worship for emperors of various dynasties of China as well as a splendid representation of ancient Chinese architectural art. These sites have been selected by the United Nations Educational, Scientific and Cultural Organization as representing the world cultural heritage. Hutong (Chinese alleys) and compound courtyards (old Beijing residential quarters) are found throughout Beijing. These streets and buildings have witnessed the ups and downs of the city and the people in past centuries and are symbolic of the life of Beijing people. Few cities have the unique historical charm of Beijing. Its wide thoroughfares, magnificent gate tower and memorial arch, and grand palaces all speak to the extensive history of this city. It also stands as a symbol of China's grandeur, history, culture and mystery. Beijing is also an approachable and visitor-friendly city.

Changes have been taking place day-by-day in Beijing since China's reform and opening to the outside world. As summarized in a popular saying, Beijing is growing taller with more and more skyscrapers while growing younger with the improving living standards and more diversified life style. This is Beijing, old and young, full of attractions. It is our sincere wish that you will make the best of your time here and we believe you will bring home more than what you expect.

xii Attractions

• Great Wall

A Chinese saying goes that He who has never been to the Great Wall is not a true man. If we laid the bricks and rocks used in the Great Wall of Ming to form a wall one meter (1.1 yard) wide and five meters (16.4 feet) high, it could circle the earth at the equator with great ease. It is such a spectacular and formidable architectural feat that anyone who comes to China should not miss it under any circumstances. The Badaling Great Wall, constructed in 1502 (during the Ming Dynasty), once served as a crucial military fortification, and is now the most impressive and representative section of the striking Great Wall. It is about 70 kilometers (43.4 miles) from the downtown area of Beijing. As Badaling was once an important military strategy point, here the wall is comparatively high and firm. It has a length of 3,741 meters (2.3 miles) and it is equipped with dense watchtowers. The wall is about 8.5 meters (27.9 feet) high and slopes inward as it rises in height. The wall is 6.5 meters (21.3 feet) wide at its base, and its rim spans about 5.7 meters (18.7 feet) across.

• Tiananmen Square

Tiananmen Square is the geographical center of Beijing City. It is the largest city square in the world, occupying an area of 440,000 square meters (about 109 acres), and able to accommodate 10,000,000 people at one time. In the center of the Square stands the Monument to the People's Heroes, which commemorates the martyrs who devoted their lives to the Chinese people. It reaches 37.94 meters (124 feet) which makes it the biggest monument in Chinese history. The body is made of hardy granite and is surrounded by white balusters. Tiananmen Tower in the south was built in 1417 during the Ming Dynasty (1368-1644). During this dynasty and the following Qing Dynasty (1644-1911) it was where proclamations were issued to the whole nation. The common people were prohibited from entering the tower, but now xiii tourists with tickets are permitted to climb it. It has five arched gates and nine principle hall columns. With the delicately carved white marbles on its base and yellow tiles on the roof, the tower is quite resplendent. Under the tower flows the limpid Jinshui River, across which seven exquisite bridges are perched, named the Golden Water Bridges.

• Beijing Lama Temple

Beijing Lama Temple is one of the largest and most important Buddhist Tibetan monasteries in the world. Construction and works in the church to unite the Han Chinese and Tibetan styles. This story is as follows. Construction work at the Yong He Gong Lama Temple began in 1694 during the Qing Dynasty. Initially, he served as official residence for court eunuchs. It was then converted to a court Prince Yong (Yin Zhen), son of Emperor Kangxi and Emperor Yongzheng himself a future. After Yongzheng ascension to the throne in 1722, half of the building was converted into a monastery, a monastery for monks of Tibetan Buddhism. The other half was left of the Imperial Palace. After Yongzheng’s death in 1735, his coffin was placed in the temple. Emperor Qianlong, Yongzheng’s successor, gave the temple imperial status is indicated with its turquoise tiles replaced by yellow tiles, which were reserved for the emperor. Subsequently, the monastery became a residence for large numbers of Tibetan Buddhist monks from Mongolia and Tibet, and so Yonghe Monastery has become a national center of Lama administration. The temple is said to have survived the Cultural Revolution because of the intervention of Prime Minister Zhou Enlai. It was opened to the public in 1981.

• Summer Palace

The Summer Palace, Yiheyuan in Chinese, is the most celebrated imperial garden in China. The garden came into existence early in the 1750s and had once been a summer resort for the emperors. It is acclaimed as a museum of gardens in China, for a visit to this garden bestow on sightseers a glimpse of representative scenes all over China.

xiv Conference Information

Conference Registration

A conference registration desk will be opened online from UTC+8 (Beijing Time): 13:30 on October 13 to 11:00 on October 16 in Online Conference Room 1.

October 13, 2020 13:30~18:30 (Online Conference Room 1) October 14, 2020 07:30~18:30 (Online Conference Room 1) October 15, 2020 08:00~18:00 (Online Conference Room 1) October 16, 2020 08:00~11:00 (Online Conference Room 1)

Social Events

The social events organized by the IEEE ICMA 2020 include the awards banquet and the farewell party, which will be held by online in Online Conference Room 1.

Awards Banquet

The Awards Banquet will be held online from UTC+8 (Beijing Time): 18:30 to 21:00 on October 15, 2020 in Online Conference Room 1. All the conference participants are welcome to join this event.

Farewell Party

The Farewell Party will be held online from UTC+8 (Beijing Time): 12:15 to 13:00 on October 16, 2020 in Online Conference Room 1. All the conference participants are welcome to join this event.

xv IEEE ICMA 2020 Conference Plenary Talk 1

Intelligent Robotic System with Machine Learning

Toshio Fukuda, Ph.D.

Professor

Meijo University, Japan

5 Sout1 Chome-501 Shiogamaguchi, Tempaku Ward, Nagoya, Japan

Postcode: 486-8502, Japan

E-mail: [email protected]

http://www.mein.nagoya-u.ac.jp/

xvi Abstract:

Recent robot technology (RT) has made remarkable progress in both manufacturing and service sectors. Because of this RT advanced technology, there are growing demands to make robots work more friendly and flexible coordinated with human for service. There are many research and developing works undergoing for robot and human interaction, such as assistance and supports of human by robots in manufacturing, inspection and maintenance, entertainment, education, bio-medical applications, rehabilitation and techno-care of aged people. Robot is required to have the more flexibility and adaptation control to human behavior, more friendly robot and human interface, and estimation capability of human intention some way to make more proactive motion. There are a lot of problems to solve them with robotic sensor, actuator, control, communication and interface with human. This talk is an overview of the Multi-scale robotics, based on the Cellular Robotics System, which is the basic concept of the emergence of intelligence in the multi-scale way from Organizational Level, Distributed robotics to Biological Cell engineering and Nano- robotics, which was proposed more than 30 years ago. It consists of many elements how the system can be structured from the individual to the group/society levels in analogy with the biological system. It covers with the wide range of challenging topics, but intelligent robotics is focused in this talk: 1. Distributed autonomous robotic system 2. Cooperation and competition of the multiple robotics system 3. Individual robot level, Brachiation Robots and Multi-locomotion robots, 4. Medical robotics and simulator, 5. Micro and nano robotics system 6. Bio analysis and bio-synthesis: bio-robotics system 7. Cyborg and Bionic System 8. Other systems. Then I mainly focus on Robotics and AI and refer to applied areas for the future hybrid system to improve the quality of life of human.

Toshio Fukuda (M'83-SM'93-F'95) graduated from Waseda University, Tokyo, Japan in 1971 and received the Master of Engineering degree and the Doctor of Engineering degree both from the University of Tokyo, in 1973 and 1977, respectively. He studied at Graduate School of Yale University in 1973-1975. He joined the National Mechanical Engineering Laboratory in Japan in 1977, the Science University of Tokyo in 1982, and then joined Department of Mechanical Engineering, Nagoya University, Japan in 1989. He worked at University of Stuttgart, as Humboldt Fellow in 1979-1981. He is Professor Emeritus of Nagoya University and Professor of Meijo University and Waseda University as well as Beijing Institute of Technology. He has been working as

xvii Professor of Shenyang University of Technology, Suzhou University, Institute of Automation, Chinese Academy of Science, Russell Springer Chaired Professor at UC Berkeley, Seoul National University, Advisory Professor of Industrial Technological Research Institute in Taiwan and etc. He is a Foreign member of Chinese Academy of Sciences (2017). He is mainly engaging in the research fields of intelligent robotic system, micro and nano robotics, bio-robotic system, and technical diagnosis and error recovery system.

He was the President of IEEE Robotics and Automation Society (1998-1999), Director of the IEEE Division X, Systems and Control (2001-2002), the Founding President of IEEE Nanotechnology Council (2002-2005), Region 10 Director (2013-2014), Director of Division X, Systems and Control (2017-2018) and IEEE President (2020). He was Editor- in-Chief of IEEE/ASME Trans. Mechatronics (2000-2002).

He was the Founding General Chair of IEEE International Conference on Intelligent Robots and Systems (IROS) held in Tokyo (1988). He was Founding Chair of the IEEE Conference on Nanotechnology(2001), IEEE Workshop on Robot and Human Communication (1994), IEEE Workshop on Advanced Robotics Technology and Social Impacts (ARSO, 2005), Founding Chair of the IEEE Workshop on System Integration International (SII, 2008), Founding Chair of the International Symposium on Micro-Nano Mechatronics and Human Science (MHS, 1990-2012), IEEE Conference on Cyborg and Bionic Systems(2017), IEEE Conference on Intelligence and Safety of Robots (2018).

He has received many awards such as IEEE Eugene Mittelmann Achievement Award (1997), IEEE Third Millennium Medal (2000) , Humboldt Research Prize (2003), IEEE Robotics and Automation Pioneer Award (2004), IEEE Transaction Automation Science and Engineering Googol Best New Application Paper Award (2007), George Saridis Leadership Award in Robotics and Automation (2009), IEEE Robotics and Automation Technical Field Award (2010). He received the IROS Harashima Award for Innovative Technologies (2011), Friendship Award of Liaoning Province PR China (2012), Friendship Award from Chinese Government (2014), JSME Achievement Award (2015), IROS Distinguished Service Award (2015) and Honor of Medal with the Purple Ribbon from Japanese Government (2015). Award from Automation Foundation (2016), Chunichi Culture Award (2019).

IEEE Fellow (1995). SICE Fellow (1995). JSME Fellow (2002), RSJ Fellow (2004), VRSJ Fellow (2011) and member of Science Council of Japan (2008-2014), and Academy of Engineering of Japan (2013-).

xviii IEEE ICMA 2020 Conference Plenary Talk 2

Microrobotics and Nanomedicine: Future Directions in Medical Robotics

Bradley Nelson, Ph.D. 㻌

Professor, ETH Zurich

Head of Department of Mechanical and Process Eng.

Tannen Strasse 3, 8092 Zurich, Switzerland

ETH Zurich, Switzerland

xix

Abstract:

While the futuristic vision of micro and nanorobotics is of intelligent machines that navigate throughout our bodies searching for and destroying disease, we have a long way to go to get there. Progress is being made, though, and the past decade has seen impressive advances in the fabrication, powering, and control of tiny motile devices. Much of our work focuses on creating systems for controlling micro and nanorobots as well as pursuing applications of these devices. As systems such as these enter clinical trials, and as commercial applications of this new technology are realized, radically new therapies and uses will result that have yet to be envisioned.

Brad Nelson has been the Professor of Robotics and Intelligent Systems at ETH Zürich since 2002. He has over thirty years of experience in the field of robotics and has received a number of awards in the fields of robotics, nanotechnology, and biomedicine. He serves on the advisory boards of a number of academic departments and research institutes across

North America, Europe, and Asia and is on the editorial boards of several academic journals. Prof. Nelson is the Department Head of Mechanical and Process Engineering at

ETH and has been the Chairman of the ETH Electron Microscopy Center and a member of the Research Council of the Swiss National Science Foundation. He also serves on boards of three Swiss companies. Before moving to Europe, Prof. Nelson worked as an engineer at

Honeywell and Motorola and served as a United States Peace Corps Volunteer in Botswana,

Africa. He has also been a professor at the University of Minnesota and the University of

Illinois at Chicago.

xx IEEE ICMA 2020 Conference Plenary Talk 3

Model Driven Mechatronics in Medical Robot Design

Blake Hannaford, Ph.D.

Professor, Department of Electrical & Computer Engineering

Director, UW Biorobotics Lab

University of Washington

Seattle WA USA

https://www.ece.uw.edu/people/blake-hannaford

xxi

Abstract:

Development of robotic technology for medical application is a complex endeavor. Engineers with little or no formal medical training must understand highly evolved medical requirements as well as the state of the art in robotics engineering. The standard for accepting a new medical innovation prior to use in the clinic is extremely rigorous. In light of these challenges, even in the more relaxed constraints of research devices, model driven design and control engineering is required to supplement highly evolved and rigorously trained human skills. Lacking a general theory of models in medical robotics, this talk will introduce the use of models in medical robot system control and design through examples from the work of our students and collaborators.

Blake Hannaford received the B.S. degree in Engineering and Applied Science from Yale University in 1977, and the M.S. and Ph.D. degrees in Electrical Engineering from the University of California, Berkeley. From 1986 to 1989 he worked on the remote control of robot manipulators in the Man-Machine Systems Group in the Automated Systems Section of the NASA Jet Propulsion Laboratory, Caltech and supervised that group from 1988 to 1989. Since September 1989, he has been at the University of Washington in Seattle. He was awarded the National Science Foundation's Presidential Young Investigator Award, the Early Career Achievement Award from the IEEE Engineering in Medicine and Biology Society, and was named IEEE Fellow in 2005. He was at Google X / Google Life Sciences from April 2014 to December 2015. His currently active research interests include surgical robotics, surgical skill modeling, and haptic interfaces.

Blake Hannaford, Ph.D., is Director of Technical Programs and the University of Washington's, Global Innovation Exchange (GIX), and Professor of Electrical & Computer Engineering, Adjunct Professor of Bioengineering, Mechanical Engineering, and Surgery at the University of Washington.

xxii

IEEE ICMA 2020 Conference Plenary Talk 4

Soft Robotics Leading to E-kagen Science and Technology

Koichi SUZUMORI, Ph.D.

Professor, Tokyo Institute of Technology

http://www.robotics.mech.e.titech.ac.jp

Abstract:

In this presentation, I will discuss three topics on soft robotics. First, I will talk about my works on soft robots. Since 1986, I have been developing various types of soft actuators; they include pneumatic rubber actuators, thin artificial muscles, functional rubber surfaces, and ion conductive polymer actuators. I will be discussing them and their applications to medical robots, soft power support suits, musculo-skeletal robots, and Giacometti robots.

xxiii

Then, I will talk about the MEXT KAKENHI project on soft robots in Japan, which was initiated in 2018 with a budget of 1.2 billion yen and a research period of five years. It is an interdisciplinary project, in which approximately 100 researchers and students participate from various fields such as mechanical engineering, electrical engineering, computer science, biology, zoology, medicine and etc. I will talk about several examples of our recent progress. At the end of my talk, I’d like to discuss about the significance of soft robotics in the history of science and technology. I think soft robotics is a value changer. While traditional robotics has been seeking for power, accuracy, and certainty, soft robotics seems to accept and utilize the contracting features, uncertainty and fuzziness, for example to accomplish jobs easily which traditional robotics has difficulty to do. Similar trends can be found recently not only in robotics but also in other fields such as material science, computer science, and industries. I will discuss this trend with the help of a Japanese word “E-kagen”, which has two contrasting meanings and I will conclude that soft robotics can be a pioneer of this trend in science and technology. Koichi Suzumori received his Ph.D. degree in mechanical engineering from Yokohama National University in 1990. He worked for Toshiba R&D Center from 1984 to 2001, and for Micromachine Center, Tokyo, from 1999 to 2001. He was then a Professor at Okayama University from 2001 to 2014. Since 2014, he has been a Professor at Tokyo Institute of Technology. He has developed various types of new actuators and applied them to new robots including soft robots, micro robots, and tough robots. He established a start-up venture company, s-muscle Co., Ltd., in 2016, which puts his soft thin artificial muscles into practical uses. He is a project leader of the MEXT KAKENHI project on soft robots. More information can be obtained in http://www.robotics.mech.e.titech.ac.jp/home.html http://softrobot.jp/en/ https://www.s-muscle.com/ https://www.youtube.com/channel/UCOz02yvsxufQPe1vWmlANAA/videos https://www.youtube.com/channel/UC1-ZyZ_G4Z5R3KUsYNDZYLQ/videos

xxiv IEEE ICMA 2020 Program at a Glance You can click Here to download the zoom software You can click the blue hyperlink to get access to the online conference room October 13-16, 2020

Online Conference Beijing Friendship Hotel Beijing, China Beijing Time Tuesday, October 13, 2020 UTC+8: 13:30 - 18:30 Registration Desk Open (Online in Conference Room 1) UTC+8: 16:00 - 17:00 Plenary Talk #1 (Prof. Toshio Fukuda) (Online Conference Room 1) Beijing Time Wednesday, October 14, 2020 UTC+8: 08:30 - 09:00 Opening Ceremony (Online in Conference Room 1) UTC+8: 09:00 - 09:50 Plenary Talk #2 (Dr. Bradley Nelson) (Online Conference Room 1) UTC+8: 09:50 - 10:40 Plenary Talk #3 (Dr. Blake Hannaford) (Online Conference Room 1) UTC+8: 10:40 - 11:00 Morning Break UTC+8: 11:00 - 12:00 Technical Sessions WA1 (Poster Session) (Online Conference Room 1) UTC+8: 12:00 - 13:30 Lunch Break UTC+8: 13:30 - 15:00 Technical Sessions WP1 (Online Room 1 -Room 6) UTC+8: 15:00 - 15:15 Afternoon Break UTC+8: 15:15 - 16:45 Technical Sessions WP2 (Online Room 1 -Room 6) UTC+8: 17:00 - 18:30 Technical Sessions WP3 (Online Room 1 -Room 6)

Beijing Time Thursday, October 15, 2020 UTC+8: 08:30 - 09:30 Plenary Talk #4 (Dr. Koichi SUZUMORI) (Online Conference Room 1) UTC+8: 09:30 - 11:00 Technical Sessions TA1 (Online Room 1 -Room 6) UTC+8: 11:00 - 11:15 Morning Break UTC+8: 11:15 - 12:15 Technical Sessions TA2 (Online Room 1 -Room 6) UTC+8: 12:15 - 13:30 Lunch Break UTC+8: 13:30 - 15:00 Technical Sessions TP1 (Online Room 1 -Room 6) UTC+8: 15:00 - 15:30 Afternoon Break UTC+8: 15:30 - 17:00 Technical Sessions TP2 (Online Room 1 -Room 6) UTC+8: 18:30 - 21:00 Award Banquet (Online in Conference Room 1)

Beijing Time Friday, October 16, 2020 UTC+8: 08:30 - 10:00 Technical Sessions FA1 (Online Room 1 -Room 6) UTC+8: 10:15 - 10:30 Morning Break UTC+8: 10:30 - 12:00 Technical Sessions FA2 (Online Room 1 -Room 6) UTC+8: 12:15 - 13:00 Farewell Party (Online in Conference Room 1) * 15 minutes (Speech: 12 minutes, Q&A:3 minutes) are scheduled for oral presentation including discussions for each paper *30 minutes (core time) are scheduled for poster presentation

xxv

IEEE ICMA 2020 Technical Program, Tuesday, October 13, 2020 Room Room 1 Room 2 Room 3 Room 4 Room 5 Room 6 Beijing Time UTC+8: 13:30 - 18:30 Registration Desk Open (Online in Conf. Room 1) UTC+8: 16:00 - 17:00 Plenary Talk #1 (Prof. Toshio Fukuda) (Online Conf. Room 1) IEEE ICMA 2020 Technical Program, Wednesday, October 14, 2020 Room Room 1 Room 2 Room 3 Room 4 Room 5 Room 6 Beijing Time UTC+8: 8:30 - 9:00 Opening Ceremony (Online Conf. Room 1) UTC+8: 9:00 - 9:50 Plenary Talk #2 (Dr. Bradley Nelson) (Online Conf. Room 1) UTC+8: 9:50 - 10:40 Plenary Talk #3 (Dr. Blake Hannaford) (Online Conf. Room 1) UTC+8: 10:40 - 11:00 Morning Break UTC+8: 11:00 - 12:00 Technical Sessions WA1 (Poster Session) (Online Conf. Room 1) UTC+8: 12:00 - 13:30 Lunch Break WP1-1 WP1-2 WP1-3 WP1-4 WP1-5 WP1-6 UTC+8: 13:30 - 15:00 Intelligent Biomedical Instrument Intelligent Mechatronics and Industrial, Manufacturing Process Vision System, Robotic Vision (I) Space Robotics and Mechatronics Biomimetic Underwater Robots Technology Application and Automation (I) UTC+8: 15:00 - 15:15 Afternoon Break WP2-1 WP2-2 WP2-3 WP2-4 WP2-5 WP2-6 xxvi UTC+8: 15:15 - 16:45 Robot Navigation and Control Manipulator Control and Industrial, Manufacturing Process Elements, Structures, and Biomimetic Measurement and Vision System, Robotic Vision (II) Algorithm (I) Manipulation (I) and Automation (II) Mechanisms (I) Control in Robotics WP3-1 WP3-2 WP3-3 WP3-4 WP3-5 WP3-6 UTC+8: 17:00 - 18:30 Robot Navigation and Control Manipulator Control and Industrial, Manufacturing Process Elements, Structures, and Micro and Nano Systems Biomimetic Systems Algorithm (II) Manipulation (II) and Automation (III) Mechanisms (II) IEEE ICMA 2020 Technical Program, Thursday, October 15, 2020 Room Room 1 Room 2 Room 3 Room 4 Room 5 Room 6 Beijing Time UTC+8: 8:30 - 9:30 Plenary Talk #4 (Dr. Koichi SUZUMORI) (Online Conf. Room 1) TA1-1 TA1-2 TA1-3 TA1-4 TA1-5 TA1-6 UTC+8: 9:30 - 11:00 Modeling, Simulation Techniques Cognitive Science and Neural Medical, Biomedical and Control Theory and Application (I) Signal and Image Processing (I) Mobile Robot System (I) and Methodologies (I) Engineering Methods (I) Rehabilitation Systems (I) UTC+8: 11:00 - 11:15 Morning Break TA2-1 TA2-2 TA2-3 TA2-4 TA2-5 TA2-6 UTC+8: 11:15 - 12:15 Control Theory and Application Modeling, Simulation Techniques Cognitive Science and Neural Medical, Biomedical and Signal and Image Processing (II) Mobile Robot System (II) (II) and Methodologies (II) Engineering Methods (II) Rehabilitation Systems (II) UTC+8: 12:15 - 13:30 Lunch Break TP1-1 TP1-2 TP1-3 TP1-4 TP1-5 TP1-6 UTC+8: 13:30 - 15:00 Control Theory and Application Modeling, Simulation Techniques Cognitive Science and Neural Medical Robots for Minimal Signal and Image Processing (III) Mobile Robot System (III) (III) and Methodologies (III) Engineering Methods (III) Invasive Surgery (I) UTC+8: 15:00 - 15:30 Afternoon Break TP2-1 TP2-2 TP2-3 TP2-4 TP2-5 TP2-6 UTC+8: 15:30 - 17:00 Control Theory and Application Modeling, Simulation Techniques Cognitive Science and Neural Medical Robots for Minimal Signal and Image Processing (IV) Mobile Robot System (IV) (IV) and Methodologies (IV) Engineering Methods (IV) Invasive Surgery (II) UTC+8: 18:30 - 21:00 Award Banquet (Online in Conf. Room 1) IEEE ICMA 2020 Technical Program, Friday, October 16, 2020 Room Room 1 Room 2 Room 3 Room 4 Room 5 Room 6 Beijing Time FA1-1 FA1-2 FA1-3 FA1-4 FA1-5 FA1-6 #Pending (Author who missed his/her oral UTC+8: 8:30 - 10:00 Neuro, Fuzzy, and Intelligent Rotor Dynamics, Vibration Surgical Robot System Digital Design and Manufacture Humanoid Robots presentation due to network Control (I) Analysis and Vibration Control problems will be arranged in this session.) UTC+8: 10:15 - 10:30 Morning Break FA2-1 FA2-2 FA2-3 FA2-4 FA2-5 FA2-6 #Pending (Author who missed his/her oral UTC+8: 10:30 - 12:00 Neuro, Fuzzy, and Intelligent Sensor Networks, Distributed Human-System Interaction and Control Theory and Application Soft Robotics presentation due to network Control (II) Sensor Systems Interface (V) problems will be arranged in this session.) UTC+8: 12:15 - 13:00 Farewell Party (Online in Conf. Room 1) xxv i

IEEE ICMA 2020 Online Conference Room Information Meeting Room Meeting ID Passcode URL IEEE ICMA 2020 https://us02web.zoom.us/j/84615945401?pwd=WEpQN3NyZFpNWWR 846 1594 5401 icma2020 Online Conforence Room 1 JNkNQaXEvbFJPZz09 IEEE ICMA 2020 https://us02web.zoom.us/j/84943657558?pwd=Q3lBZ05DYktRVG10ME 849 4365 7558 icma2020 Online Conforence Room 2 FOOUVPejY5QT09 IEEE ICMA 2020 https://us02web.zoom.us/j/87280341853?pwd=QWxGckpKUjlqbVlVY1 872 8034 1853 icma2020 Online Conforence Room 3 pKVFRvdEJOUT09 IEEE ICMA 2020 https://us02web.zoom.us/j/83180902952?pwd=cVo2TjAxSFhkd3FzRlhr 831 8090 2952 icma2020 Online Conforence Room 4 YnRRblVLdz09 IEEE ICMA 2020 https://us02web.zoom.us/j/81969274417?pwd=VUJQTFJrTXJWd2ZxdE 819 6927 4417 icma2020 Online Conforence Room 5 ZUMFNUbjJrQT09 IEEE ICMA 2020 https://us02web.zoom.us/j/82091055540?pwd=cG5FL3c4UHBmVHh2S 820 9105 5540 icma2020 Online Conforence Room 6 mRXTklKZC9Fdz09

Wednesday October 14, 2020

Morning Sessions

WA1-P Poster Session (Intelligent Mechatronics and Automation)

Wednesday October 14, 2020

Afternoon Sessions

WP1-1 Intelligent Biomedical Instrument Technology WP1-2 Intelligent Mechatronics and Application WP1-3 Industrial, Manufacturing Process and Automation (I) WP1-4 Vision System, Robotic Vision (I) WP1-5 Space Robotics and Mechatronics WP1-6 Biomimetic Underwater Robots WP2-1 Robot Navigation and Control Algorithm (I) WP2-2 Manipulator Control and Manipulation (I) WP2-3 Industrial, Manufacturing Process and Automation (II) WP2-4 Vision System, Robotic Vision (II) WP2-5 Elements, Structures, and Mechanisms (I) WP2-6 Biomimetic Measurement and Control in Robotics WP3-1 Robot Navigation and Control Algorithm (II) WP3-2 Manipulator Control and Manipulation (II) WP3-3 Industrial, Manufacturing Process and Automation (III) WP3-4 Micro and Nano Systems WP3-5 Elements, Structures, and Mechanisms (II) WP3-6 Biomimetic Systems

IEEE ICMA 2020 Conference Digest WA1-P: Poster Session (Intelligent Mechatronics and Automation)

Session Chairs: Liwei Shi, Beijing Institute of Technology Xiaoliang Jin, Kagawa University Online Conference Room 1, UTC+8(Beijng Time): 11:00-12:00, Wednesday, 14 October 2020

WA1-P(1) 11:00-12:00 WA1-P(2) 11:00-12:00

Research on Mobile Robot Path Planning Based A Novel Misoperation Preventing Device for the on Deep Reinforcement Learning Interventional Surgical Robot XiuFen Ye, Shuo Zhang Cheng Yang 1, Shuxiang Guo1, 2 * College of Intelligent Systems Science and Engineering., Harbin Engineering University 1. Key Laboratory of Convergence Medical Engineering System and Healthcare Technology, Harbin, Heilongjiang Province , China the Ministry of Industry and Information Technology, School of Automation, Beijing Institute of Technology, No.5, Zhongguancun South Street, Haidian District, Beijing, China. • Path planning based on deep 2. Faculty of Engineering, Kagawa University, 2217-20 Hayashi-cho, Takamatsu, Kagawa, Japan. reinforcement learning.  Based on previous research in our laboratory, • Increased generalization ability of this paper proposes a novel misoperation agents by adding adaptive gaussian preventing device for the robot operation safety during interventional surgery and evaluates its noise. gripping performance through experiments. • Redesigned the reward functions  Compared with our previous research, the novel based on the curiosity mechanism. misoperation preventing device is less affected by voltage and can be operated more accurately. • Added the memory units for agents Experiments show that the novel misoperation to make the exploration process The novel gripping device Mobile Robot Path Planning gripping device can complete the clamping of more efficient. guidewire by applying a pressure of more than structure diagram. 17.5N.

WA1-P(3) 11:00-12:00 WA1-P(4) 11:00-12:00 Research and Analysis of Drilling String Vibration Signals SVM Classification for Novel Time Domain IMU Bilin Li,Yi Gao, and Ruibing Li and EMG fused features for control of 6-DOF School of Electronic Engineering.,Xi'an Shiyou University Xi'an, China industrial robot • This paper focuses on the generation Haider Ali, Wang Yanen mechanism, performance Department of Mechanical Engineering., Nortthwestern Polytechnical University phenomenon, negative effects, Xian, China judgment and improvement measures of three forms of drilling tool vibration: longitudinal vibration,torsional vibration and lateral vibration. The three vibration Fig 1.Overall layout of the Project signal research methods: time • Gestures were recognized using EMG and IMU sensors. domain method , spectrum analysis method and wavelet transform Torsional vibration Longitudinal vibration Lateral vibration • The gestures were then used to control the industrial robot method are also analyzed. Three forms of vibration whilst avoiding the singularities in the path intended.

WA1-P(5) 11:00-12:00 WA1-P(6) 11:00-12:00

An Implementation Strategy of Simplified Research and Application Based on the Soft Restriction Design of Product Design Maximum Power Tracking Algorithm Ying Zhang, Xiaodong Yuan, Yunliang Wang, Shengpu Zhang Design Basic Teaching Department , Tianjin University of Technology ,Tianjin 300384 Jilin Animation Institute, Tianjin, China Changchun, Jilin, China • 1.Present a new design method and theory to make the design of the product more "safety", • Mathematical model of photovoltaic "efficient", " reasonable " for users. cell. • Output characteristics of photovoltaic • 2.Regulate the user's bad behavior, for the user's reasonable life, long-term health benefits to cell. bring help and positive influence, and reached • Maximum power tracking algorithm. the advanced human engineering standards with proposed design method. • Simulation circuit. • Simulation analysis. The Voltage Curves Of • 3.The design idea and method of the Soft Two Algorithms Restriction Design have lots of benefits on rehabilitation devices design for patients’ safety and comfort. Use software for product modeling

IEEE ICMA 2020 Conference Digest WA1-P: Poster Session (Intelligent Mechatronics and Automation)

Session Chairs: Liwei Shi, Beijing Institute of Technology Xiaoliang Jin, Kagawa University Online Conference Room 1, UTC+8(Beijng Time): 11:00-12:00, Wednesday, 14 October 2020

WA1-P(7) 11:00-12:00 WA1-P(8) 11:00-12:00

Fault Diagnosis of Asynchronous Motor Based on Envelope Spectrum and Step Frequency Division Experimental study on a high-speed spline-quill- Gao Ya, Li Bo, Zhu Qinling, Gao Yi shaft coupling-rotor-bearing system Electronic information engineering college, Xi’an Technological University Sheng Feng, Jiangang Zhu, Haipeng Geng, Baisong Yang,Lie Yu Xi’an, China State Key Laboratory for Strength and Vibration of Mechanical Structures, • A composite fault separation method Xi’an Jiao tong University, Xi’an, Shaanxi Province, China is proposed to solve the problem of multiple fault frequencies in the • Install the test bench. current spectrum of asynchronous • The gearbox is idling without motors. connecting the test rotor. • Mean square noise reduction. • Test the rotor system. • Spectrum analysis, these include • System critical speed based on low-frequency energy spectrum experimental analysis. analysis and low-frequency energy • System critical velocity calculated spectrum analysis. The Faulty Motors by finite element method. picture of the test rotor

WA1-P(9) 11:00-12:00 WA1-P(10) 11:00-12:00

Experimental study on a 100kW high speed Application of Hybrid Energy Storage Device in permanent magnet synchronous motor High Power Direct-current Power Supply Yue Liu, Chunjie Wang, Peng Chen and Jinliang Yin supported by gas foil bearings School of Electrical and Electronic Engineering Tianjin University of Technology Sheng Feng, Xiaobo Dai, Baisong Yang, Jiale Tian, Hao Lin，Lie Yu Tianjin, China State Key Laboratory for Strength and Vibration of Mechanical Structures, Xi’an Jiaotong • A design scheme of high-power DC University, Xi’an, Shaanxi Province, China power supply with super capacitor and battery as energy storage is Energy Storage Unit DC Bus • Set up an experimental platform. proposed. • The turbo blower accelerates from • Hybrid energy storage system uses High power Super DC circuit active topology. Battery DC power zero speed to 15000rpm and capacitor breaker supply measures the power, voltage, current • Theoretical and simulation studies and power factor of the motor. on energy management and • After the experiment, the load side discharge control strategies were Overall block diagram gas foil thrust bearing wear and The motor rotor conducted to verify the effectiveness of the system impeller vibration. and feasibility of the scheme.

WA1-P(11) 11:00-12:00 WA1-P(12) 11:00-12:00 A Structural Design of the Assembling Multi- A Review of Fault Diagnosis Methods of Robot functional Capsule Robot Driving Module Joint Servo System Hong Wu, Jiasi Feng Lining Zhang, Shuxiang Guo, Dongdong Bu, Youchun Ma School of Electronic Engineering, Heilongjiang University Key Laboratory of Convergence Biomedical Engineering System and Healthcare Technology, Harbin, China The Ministry of Industry and Information Technology, School of Life Science, Beijing Institute of Technology • The mechanism of motor fault and Beijing, China drive fault in joint servo system is • The assembled multifunctional analyzed. capsule robot is composed of two modules • The fault diagnosis methods proposed in recent years are • When the robot works, the driver summarized and their advantages module and the required functional and disadvantages are discussed. modules must be assembled together. • the future development trend of joint servo system fault diagnosis • The robot driving module can The structure of the assembling technology is given. move stably in the pipe by rotating multi-functional capsule robot. Fault type of PMSM magnetic field.

2 IEEE ICMA 2020 Conference Digest WA1-P: Poster Session (Intelligent Mechatronics and Automation)

Session Chairs: Liwei Shi, Beijing Institute of Technology Xiaoliang Jin, Kagawa University Online Conference Room 1, UTC+8(Beijng Time): 11:00-12:00, Wednesday, 14 October 2020

WA1-P(13) 11:00-12:00 WA1-P(14) 11:00-12:00 Classification of Subject-Independent Motor An Incremental Feature Extraction Method for Imagery EEG Based on Weld Surface Defect Recognition Jun Liu, Keguang Ren, Xiaofeng Wang, and Weimin Ge Relevance Vector Machine 1.Tianjin Key Laboratory for Advanced Mechatronic System Design and Intelligent Control, Enzeng Dong, Chao Yue, and Shengzhi Du Tianjin University of Technology, Tianjin 300384, China. School of Electrical and Electronic Engineering.,Tianjin University of Technology 2. National Demonstration Centre for Experimental Mechanical and Electrical Engineering TianJin, China Education, Tianjin University of Technology, Tianjin 300384, China.  To meet the requirements of online feature extraction for weld defect q×q Row Similar Scatter Matrix q×1 Row Estimation • Introduces the basic theory of 1×q Vector recognition, an Incremental 2DPCA p×q Image p×q Image Approximation

… p Suject-Independent MI-BCI Row algorithm (I2DPCA) is proposed. Centralization p×1 Vector

Column Approximation • Combine CSP algorithm with • To extract more structure features … RVM classifier and reduce the dimension of weld q p×p Column Similar Scatter Matrix p×1 Column Estimation • Offline experiment images, an Incremental row-column • Result analysis Bidirectional PCA (IBDPCA) Block diagram of the Model building flowchart algorithm is presented based on IBDPCA algorithm I2DPCA.

WA1-P(15) 11:00-12:00 WA1-P(16) 11:00-12:00 Design and Implementation of Ansys-based Optimal Design of Cable Fault Identification Scraper Conveyor Permanent Magnet Frequency Based on the Combination of BP and GSA Zeyu Huang,Sen Wang,Jieqiu Bao,Hongkui Yan,Yan Bao,Di Tang,Yujia Li Conversion Integrated Machine Shenyang Institute of Engineering, Lei Wang,Jia Liu,Xinming Wang,Yuqin Zhu,Sen Wang,Zeyu Huang,Yujia Li,Di tang Shenyang,China Shenyang Research Institute of Coal Science Group Co.LTD,Key Laboratory of Coal • This article first introduces the Mine Safety Technology,Shenyang Institute of Engineering gravitational search algorithm and then Shenyang,China proposes an improved gravitational • Introducing the drive system of the scraper search algorithm—gravity attenuation conveyor. gravitational search method. • Designing the permanent magnet motor and • The cable is simulated and modeled. use RMxprt and Maxwell to simulate the • The gravity attenuation algorithm and BP three-dimensional model of the motor. neural network are combined to • Using ANSYS finite element analysis construct a new hybrid algorithm, which software to accurately calculate the motor is applied to cable fault diagnosis and loss value driven by variable frequency Scraper Conveyor Cable analysis. power supply.

WA1-P(17) 11:00-12:00 WA1-P(18) 11:00-12:00 Pressure optimal control model of high-pressure PSO-BP Neural Network-Based Gelatin Hollow tubing based on fourth-order Runge-Kutta Capsule Color Matching Model Qi Wang, Liguo Shuai, Xianze Dai, Huiling Chen method Shenzhen Research Institute, Southeast University Fujun Zhang, and Chenjian Mao Shenzhen, China Department of Automobile Engineering Chongqing Industry Polytechnic College • The experiment of gelatin color Department of Rail Transit and Aviation Service Chongqing Industry Polytechnic College matching was carried out . • This paper proposes two high-pressure tubing pressure control strategies based on the fourth- • MATLAB image processing toolbox order Runge-Kutta method. was utilized to conduct image filtering and image segmentation. • Control of tubing considers the effects of oil viscosity and elastic modulus on pressure. • The BP neural network’s initial parameters were optimized by the • Pressure stabilization control is by adjusting the particle swarm optimization opening and closing time of the oil pump at the algorithm, and the training speed inlet of the tubing. Gelatin solution and prediction accuracy have been • In boosting control, an approximate improved. relationship of the pressure and time is obtained.

IEEE ICMA 2020 Conference Digest WA1-P: Poster Session (Intelligent Mechatronics and Automation)

Session Chairs: Liwei Shi, Beijing Institute of Technology Xiaoliang Jin, Kagawa University Online Conference Room 1, UTC+8(Beijng Time): 11:00-12:00, Wednesday, 14 October 2020

WA1-P(19) 11:00-12:00 WA1-P(20) 11:00-12:00 Satellite image small target application based on Review on Development Status and Key deep segmented residual neural netwrok Technologies of Surgical Robots Zikang Wei, Yunqing Liu Yingwei Guo1, Yingjian Yang1, Mengting Feng1, Fengqiu Cao1, Hanhui Wu1, and Yan Kang1,2 School of Electrical and information engineering, Changchun University of Science and 1College of Medicine and Biological Information Engineering, Northeastern University Technology, Jilin, China Shenyang, Liaoning Province, China • Use Mish activation function 2College of Heath Science and Environmental Engineering, Shenzhen Technology University instead of ReLU activation Shenzhen, Guangdong Province, China function, which reduce the death • The development status of surgical and loss of the neurons during the robots. network training. • From three aspects of the structure • Use two residual layers divide the design, the virtual simulation system network into two parts, make the and the control factor, analyzeing residual loss achieve stable faster the key technology of surgical and make network performance robots. more better. • Predicting the future development of Raven Robot Structure of deep segmented • The super-resolution effect of the surgical robots. (Harbin Institute of Technology) residual neural network satellite image small target is better

WA1-P(21) 11:00-12:00 WA1-P(22) 11:00-12:00 Trajectory Tracking Control Based on RBF Neural Application of Chaotic Time Series Phase Space Network of The Lower Limb Rehabilitation Robot Reconstruction on Arc Fault Detecting Jia Shi1,2, Linsen Xu1, Gaoxin Cheng1,2 ,Jiajun Xu1,2, Shouqi Chen1,2 and Xingcan Liang1,2 Jieqiu Bao,Sen wang,Yi Zhang,Hongkui Yan,Huan Wang,Jinyu Li,Siqi Liu 1University of Science and Technology of China, Hefei City, Anhui Province, China College of automation, Shenyang Institute of Engineering 2Institute of Advanced Manufacturing Technology Hefei Institutes of Physical Science, CAS, Hui Shenyang, China Hong Building, Changzhou City, Jiangsu Province, China • The dynamic model of the lower limb • This paper offers an arc detection rehabilitation robot is established and - Desired + e PID τ Actual J -1 Robot J methods based the characteristics of trajectory Controller trajectory the robot joint trajectory tracking + electric arc current and chaos theory. - control scheme is built. RBFNN • This paper determines the Controller • In order to prevent the dynamic parameter reconstruction plan of the fault model uncertainties and lack of good current arc that embedding RBF adjusting PID control dimension “d” and time delay“τ”. adjustment ability, RBF-PID control block diagram is designed ,which can speed up the • Using MATLAB achieves the adjustment of parameters and make calculation methods of the the lower limb movement of the robot correlation dimension. Phase space chart more compliant.

WA1-P(23) 11:00-12:00 WA1-P(24) 11:00-12:00 Kinematics and Workspace Analysis of a New 3-DOF Workspace Analysis of 2-RPU&2-SPS Spatial Parallel Parallel Robot Mechanism Pengfei Gao1, Bin Li1, Xiaodong Wen2 and Yuebo Wang2 Xuehao Meng1, Bin Li1, Yuan Zhang2 and Shuo Xu2 1.Department of Mechanical Engineering, Tianjin University of Technology, Binshui Xidao, 1.Department of Mechanical Engineering, Tianjin University of Technology, Binshui Xidao, Xiqing District, Tianjin, China Xiqing District, Tianjin, China 2. Tianjin Key Laboratory of Aerospace Intelligent, Equipment Technology, Tianjin, China 2. Tianjin Key Laboratory of Aerospace Intelligent, Equipment Technology, Tianjin, China • This paper proposes a new type of 4- • This paper proposes a new type of 3-DOF DOF parallel mechanism. parallel mechanism. • Analyzing the degree of freedom of • The degree of freedom and properties of the the new mechanism and establishing mechanism is verified by using the helical a kinematic model. theory. • Solving the inverse kinematics of • The positive kinematics of the mechanism the mechanism and analyzing the is solved. constraints of the working space of • Solving the position workspace and posture the mechanism. The 2-RPU&2-SPS workspace of the parallel mechanism. The 2-RRU&RSR • Solving the position workspace and mechanisms mechanisms posture workspace of the mechanism.

IEEE ICMA 2020 Conference Digest WA1-P: Poster Session (Intelligent Mechatronics and Automation)

Session Chairs: Liwei Shi, Beijing Institute of Technology Xiaoliang Jin, Kagawa University Online Conference Room 1, UTC+8(Beijng Time): 11:00-12:00, Wednesday, 14 October 2020

WA1-P(25) 11:00-12:00 WA1-P(26) 11:00-12:00 The junction temperature measurement of Research on the control system of six joint insulated gate bipolar transistor based on multi- industrial robot based on CPAC Chao Dong，Bang Xie and Xin Shao layer feed -forward neural network is presented Complex System Control Theory and Application Key Laboratory., Tianjin University of Chao Dong ,Peiye Mao Technology School of Electrical and Electronic Engineering., Tianjin University of Technology Tianjin, China Tianjin, China • Kinematics modeling. • IGBT junction temperature • Hardware composition of robot prediction based on BP neural control system. network • System software design. • In this paper, neural network is • Function experiment of robot proposed to predict IGBT control system. junction temperature, and the feasibility of neural network is The neural network is compared verified with the common method（Blue: CPAC development platform • Compared with conventional neural network, red: common forecasting methods method）

WA1-P(27) 11:00-12:00 WA1-P(28) 11:00-12:00 Construction of 2.5D Map of Dynamic Environment Topology Recognition of Courts Based on Based on Information Fusion the Improved PCA Liu Bohao Du Yuhong Ren Weijia and Li Shuai Hu Jiaxiang, Zhang Bin, Long Bo, Hu Pengfei School of Mechanical Engineering，Tiangong University University of Electronic Science and Technology of China Tianjin, China Chengdu, China • Use Kinect V2 and two- • Use the principle of Principal dimensional lidar information Component Analysis to achieve fusion to build a map. time series analysis . courts1 1-A 1-B • Use YOLOv4 target detection • indicates the dimensionality 1-C users technology to eliminate reduction of time series is to transformer1 dynamic objects. compare the correlation between courts2 Power grid 2-A 2-B • Fusion of sensor information the original sequence and the 2-C users characteristic waveform sequence transformer2 using D_S evidence theory. Three-phase load balancing User energy portrait • Two-dimensional processing • Define similarity curve to amplify Abnormal power consumption detection of the point cloud information sequence differences. Eliminate dynamic object mapping User relationship recognition of Kinect v2. • Use clustering to get user relationships

WA1-P(29) 11:00-12:00 WA1-P(30) 11:00-12:00 Design of Smart Home System Based on Wechat Applet Research on Image Segmentation Method of Yunliang Wang,Feng Li Underwater Pipeline Oil Leakage Point Tianjin University of Technology Xinhua Zhao, Litao Jing and Zeshuai Du Tianjin, China Institute of Intelligent Systems and Biomedical Information, College of Automation, Harbin • WeChat Mini Program(WCMP) Engineering University software construction • EnOcean builds a wireless • Based on the characteristics communication network of the underwater target • WCMP combined with smart segmentation task, this paper home system to realize improves the segmentation environmental monitoring and network to increase the equipment control segmentation speed. • Build a server as a • The experimental results communication intermediary show that the algorithm Smart Home System improves the segmentation Improved segmentation results accuracy and speed. of network leakless pipeline

IEEE ICMA 2020 Conference Digest WA1-P: Poster Session (Intelligent Mechatronics and Automation)

Session Chairs: Liwei Shi, Beijing Institute of Technology Xiaoliang Jin, Kagawa University Online Conference Room 1, UTC+8(Beijng Time): 11:00-12:00, Wednesday, 14 October 2020

WA1-P(31) 11:00-12:00 WA1-P(32) 11:00-12:00 Research On Genetic Algorithm Of Mechanical A Multi-Feature Fusion Based Pedestrian Structure Optimization Of Weight Reduction Detection Method Enzeng Dong, Cunlei Jing,and Zufeng Zhang System Tianjin Key Laboratory For Control Theory & Applications in Complicated Systems, Tianjin 1 12，， * Huan Liu11 , Jian Li1 , Shigang Wang , Xueshan Gao University of Technology 1 GuangXi University of Science and Technology Beijing, China Tianjin, China 2 Beijing Institute of Technology Beijing, China • genetic algorithm is proposed to Mixed Feature optimize the mechanical structure. • Introduces the basic theory of Histogram features HOG feature pedestrian library • Using genetic algorithm to optimize Original detection Image mechanical structure can achieved Haar-like Rectangle Adaboost train the goal of reducing the volume and • Combine Haar feature with features weight of the structure without HOG and Adaboost decreasing the output force. • Contrast experiment Building a Hybrid Feature Database • The volume of the weight reduction The Weight Reduction System • Result analysis mechanism decreases by 48%.

WA1-P(33) 11:00-12:00 WA1-P(34) 11:00-12:00 Classification of Tire Pressure in a Semitrailer Filtering Strategies for State Estimation of Using a Convolutional Neural Network Zygimantas Ziaukas, Alexander Busch, Omniwheel Robots Mark Wielitzka, and Tobias Ortmaier Jan-Philipp Kobler Benjamin M. Dyer, Trevor Robin Smith, S. Andrew Gadsden, M. Biglarbegian University of Guelph, School of Engineering Institute of Mechatronic Systems BPW Bergische Achsen KG Leibniz Universität Hannover • Multi-tiered filtering strategies • Advanced Driver Assistance Systems were investigated for omniwheel for Passive Commercial Vehicles robots • Classification of Tire Pressure in a • Kalman filters, extended Kalman Semitrailer filter, and smooth variable • Residual Neural Network structure filter are used to improve state estimation • Time Series Classification • Accuracy of robot position and State Estimation using angle estimation is improved by multiple filtering strategies a factor of 5

WA1-P(35) 11:00-12:00 WA1-P(36) 11:00-12:00

A Robust Data Reconciliation Method Based on Design And Electromagnetic Simulation Of High Shape Parameters Speed PERMANENT Magnet Synchronous Motor Jianyun Ni, Yong Qi, Hainan Wang, Mingyang Zhao Wei Zheng, Haipeng Geng, Jiayi Yao and Hao Lv College of Electrical and Electronic Engineering, Tianjin University of Technology New Energy Equipment and Quality Engineering Research Institute Tianjin, China Xi’an Jiaotong University • Introduced the importance and Xi’an, Shanxi Province, China development of data coordination. • The design flow of permanent magnet • Propose a new data coordination motor is presented, and the design of method based on shape parameters. stator and rotor is completed. • Analyzed its main robustness and its • Ansoft Maxwell software was used for objective function and influence simulation analysis to verify the function. electromagnetic design rationality of the • The example proves the effectiveness motor. of the new robust estimation in data Petrochemical • Ansoft Maxwell analysis results were coordination. compared with analytic solutions to verify The Stator winding design the accuracy of analysis results

IEEE ICMA 2020 Conference Digest WA1-P: Poster Session (Intelligent Mechatronics and Automation)

Session Chairs: Liwei Shi, Beijing Institute of Technology Xiaoliang Jin, Kagawa University Online Conference Room 1, UTC+8(Beijng Time): 11:00-12:00, Wednesday, 14 October 2020

WA1-P(37) 11:00-12:00 WA1-P(38) 11:00-12:00 A Discrete-Time Sliding Mode Control Method Research of Helicopter Electric Tail Rotor with for Multi-Cylinder Hydraulic Press High Power-to-Weight Ratio Chao Jia, Jiaqi Wei, Enzeng Dong, Xuewei Gao, Haocheng He Qingguang Yu, Yuming Liu, Zhicheng Jiang, and Gaoxiang Long College of Electrical and Electronic Engineering, Tianjin University of Technology State Key Laboratory of Power System, Department of Electrical Engineering Tianjin, China Tsinghua University Beijing, China • In this paper, the electromotive tail • Discrete-time sliding mode control rotor system of tiltrotor has been • Main propeller • Tail rotor method for multi-cylinder hydraulic deeply studied. presses . Tail rotor reducer Tail rotor • It adopts CNN algorithm to achieve Main drive shaft • The design method of the propeller reducer the decoupling of main and tail distributed controller. Main propeller Tail rotor propeller. • Discrete-time sliding mode control • The kinematics and dynamics Motor Permanen method. Main propeller drive controller t magnet motor analysis of the electromotive tail reducer • Control allocation method. rotor are put forward. The 2D and Five-cylinder hydraulic press system 3D simulation model is designed The electromotive tail rotor structure of helicopter and implemented.

WA1-P(39) 11:00-12:00 WA1-P(40) 11:00-12:00 The Terminal Trajectory Drive and Control of a Three-Channel Soft Actuator System Design of Intelligent Control on Molding Lingyu Zhang, Junjie Zhou and Jian Zhang Production Line and Product Reliability of School of Mechanical Engineering, Beijing Institute of Technology Beijing, China Carbon Fiber Composite Minglei Xi1, Yaqiao Zhu2, Jin Wu2*, Yangyang Wang3 • Pressures coupling of a pneumatic 1 School of Mechanical Engineering, Tianjin University of Technology, Tianjin 300384, PR China drive soft actuator with three 2 Tianjin Sino-German University of Applied Sciences , Tianjin 300350, PR China paralleled channels. 3 Tianjin Windo Machine Co. , Ltd , Tianjin 301700, PR China • Based on the analysis of carbon fiber composite • Pneumatic drive model of extruding molding production line, the intelligent control and bending. system of carbon fiber composite molding • Complete PC and wireless control production line is designed. system of the soft actuator. • Experiments on processing automotive panels • Realize a circle terminal trajectory and aircraft ribs show that the product failure Intelligent Production of the soft actuator. rate of the entire production line is less than 2%. Line Coupling and Conversion of • Quality monitoring before molding can remove Soft Actuator’s Three 40%-45% of unqualified products. Channels

WA1-P(41) 11:00-12:00 WA1-P(42) 11:00-12:00 Parameters Optimization Design of 3-UPU Parallel A Novel Target Recognition System for the Mechanism based on the Spectral Norm Amphibious Robot based on Edge Computing Peng Chen, Yonggeng Wei School of Mechanical and Electrical Engineering Heilongjiang University and Neural Network Harbin, Heilongjiang Province, China Shuxiang Guo1,2 ,Handong Cheng1 ,Jian Guo1* ,Jigang Xu3* 1Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and • Taking the 3-UPU parallel mechanism as the Biomedical Robot Laboratory, Tianjin University of Technology, Tianjin, China 2 research object, an optimum design method of Intelligent Mechanical Systems Engineering Department Faculty of Engineering, Kagawa University, Kagawa, Japan parallel mechanism structure parameters based 3Qinghai Haidong,810700,China on the spectral norm is proposed. • We proposed a target recognition system based on edge computing, • The mathematical model of Jacobian matrix is The purpose is to improve the speed established. The structure parameters are of ocean rescue. optimized and simulated by MATLAB • We train the convolution neural network on the upper computer and software. use the mini yolov3 algorithm. • When the condition number is 1.58 ≤ E ≤ 1.72, • The trained data are imported into the optimized parameters meet the initial Schematic diagram the edge computing circuit board for target recognition design requirements. of mechanism The Target recognition results

IEEE ICMA 2020 Conference Digest WA1-P: Poster Session (Intelligent Mechatronics and Automation)

Session Chairs: Liwei Shi, Beijing Institute of Technology Xiaoliang Jin, Kagawa University Online Conference Room 1, UTC+8(Beijng Time): 11:00-12:00, Wednesday, 14 October 2020

WA1-P(43) 11:00-12:00 WA1-P(44) 11:00-12:00 Garlic Scale Teeth Image Recognition Based on Design of Chinese Cabbage Harvester Sobel algorithm Liwen Cao and Shuaitong Miao School of Mechanical and Electrical Engineering, Heilongjiang University Yunliang Wang, Liutao Wang, Yanjuan Wu Harbin, Heilongjiang Province, China Tianjin Key Laboratory for Control Theory and Applications in Complicated Systems, Tianjin University of Technology, Tianjin, China • Aiming at the existing cabbage production conditions in China, the • OV7670 was used for image data overall structure design of the acquisition. cabbage harvester was carried out. • Garlic seed edge detection algorithm • The working principle of the based on Sobel cabbage harvester was clarified, and the main components were modeled • Garlic seed scale tooth recognition and simulated. method based on sliding window • The designed Chinese cabbage • Identify and save feature coordinates harvester can realize the mechanization of the whole process The Chinese cabbage harvester Edge detection results of cabbage extraction, cutting, conveying.

WA1-P(45) 11:00-12:00

Semantic Segmentation of Freshwater Fish Body Based on Generative Adversarial Network Hongjun Wang,Xiaoyu Ji,Hui Zhao,Youjun Yue Tianjin University of Technology Tianjin, China

• Sample data enhancement based on C-SAGAN • Deeplabv3plus freshwater fish body semantic segmentation model • The Mean Intersection over Union (MIoU) of semantic segmentation for freshwater fish under the C- SAGAN network expanded sample set is 95%

Semantic Segmentation

IEEE ICMA 2020 Conference Digest WP1-1: Intelligent Biomedical Instrument Technology

Session Chairs: Qin Li, Beijing Institute of Technology Jin Guo, Beijing Institute of Technology Online Conference Room 1, UTC+8(Beijing Time): 13:30-15:00, Wednesday, 14 October 2020

WP1-1(1) 13:30-13:45 WP1-1(2) 13:45-14:00

Obstructive Sleep Apnea Detection Using Sleep A Novel Fuzzy Neural Network-based Rehabilitation Stage Classifying Architecture Method for the Upper Limb Rehabilitation Robotic System Shuxiang Guo1,2, Huimin Cai1 , Jian Guo1 Juan Liu, Qin Li, Yi Xin, and Xiao Lu Department of Life Science , Beijing Institute of Technology 1 Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Intelligent Robot Laboratory, Beijing, China Tianjin University of Technology, Tianjin, China • Method: Data preprocessing-Feature 2 Department of Intelligent Mechanical Systems Engineering, Faculty of Engineering, Kagawa University, Kagawa, Japan extracting-Feature selection(statistical • This paper proposes a method of collecting analysis)-Classification (RF, XGBoost, sEMG signals and interaction force to classify the LightGBM)-Feature importance rehabilitation stage and evaluate the rehabilitation process. • Results: 1. XGBoost performs better • The method solves the problem of inconvenient with the accuracy, precision, recall, f1 traditional rehabilitation process and insufficient score and AUC are 0.9130, 0.8851, number of rehabilitation doctors. 0.9506, 0.9167 and 0.9128 based on • This paper conducts a rehabilitation exercise simulation experiment based on the manual Data relationships diagram of the fuzzy selected features；2. The percentage of muscle test (MMT) scale, fits the fuzzy neural neural network. N1, the percentage of N3, one-step The Results of Classification network, and verifies its predictive performance transition pattern of N2→N1 and total for the evaluation of rehabilitation stage. number of transitions are important.

WP1-1(3) 14:00-14:15 WP1-1(4) 14:15-14:30 Magnetic Localization Technology of Capsule Path Planning of the Spherical Robot based on Robot Based on Magnetic Sensor Array Improved Ant Colony Algorithm Qiang Fu1, Xinrui Wang1, Jian Guo1, Shuxiang Guo2, Zhuocong Cai1 Jian Guo1,2 ,Xiaojie Huo1 ,Shuxiang Guo1* ,Jigang Xu3* 1Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and 1Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Biomedical Robot Laboratory,Tianjin University of Technology,Tianjin, China Intelligent Robot Laboratory, Tianjin University of Technology, Tianjin, China 2Dpartment of Intelligent Mechanical Systems Engineering Kagawa University,Takamatsu, 2Intelligent Mechanical Systems Engineering Department Faculty of Engineering, Kagawa,Japan. Kagawa University, Kagawa, Japan • This paper proposed a new 3Qinghai Haidong,810700,China • This paper presents an improved ant magnetic localizaiton algorithm colony algorithm for path planning of and compensate algorithm. spherical robots. • The proposed localizaiton • Adaptive pheromone updating rule algorithm can improve the and negative feedback mechanism are calculation speed. proposed. • It solves the problem that ant colony • The proposed compensation algorithm is easy to fall into local algorithm can further improve optimal in the early stage and has insufficient convergence in the late the positioning accuracy. Magnetic dipole model The improved ant colony algorithm looks stage. for the optimal path

WP1-1(5) 14:30-14:45 WP1-1(6) 14:45-15:00 Estimating Heart Rate during Steady-State Activities Cross-Domain Segmentation of Fundus Vessels and Transitions Based on Signal of PPG and ACC Based on Feature Space Alignment Qiang Zhang, Chuanbin Ge, Yi Xin School of Life Science, Beijing Institute of Technology Huaixin Wang, Qingshan Zeng, Lei Yang, Yanhong Liu, and Guibin Bian Beijing, China School of Electrical Engineering, Zhengzhou University, China a Institute of Automation, Chinese Academy of Sciences, China • Database: The IEEE Signal Processing Cup 2015 challenge database. • Firstly, The Wasserstein • Method: Locate the steady-state and distance is used to align two transitional state by ACC signal. Estimate different domains in the HR based on the spectrum of ACC and PPG. high-level feature space. • Result: Pearson’s correlation coefficient b • Then, a few data sets from between PPG-estimated HR and true HR target domain are used to was 0.9933, and the mean absolute error constrain the segmentation (MAE) between them was 1.3262 beats per minute (BPM). networks to improve the segmentation performance in the target domain. The System Framework a) The result of S1-TYPE01; b) The scatter of ground-truth HR and HR estimation.

IEEE ICMA 2020 Conference Digest WP1-2: Intelligent Mechatronics and Application

Session Chairs: Toshio Fukuda, Beijing Institute of Technology Chaoyang Shi, Tianjin University Online Conference Room 2, UTC+8(Beijing Time): 13:30-15:00, Wednesday, 14 October 2020

WP1-2(1) 13:30-13:45 WP1-2(2) 13:45-14:00 Visual Pencil: Design of Portable Human-Computer Interaction Teleoperation of two non-isomorphic Based on 2D Visual Tracking manipulators by Joint Ru Tong, Tianzhu Wang, and Junzhi Yu State Key Lab Management and Control for Complex Systems, Institute of Automation, CAS Space and Task Space Mapping Beijing, China Amit Shukla1, Sunil Kumar Reddy2, Hamad Karki3 School of Engineering, Indian Institute of Technology Mandi, India1 • Visual Pencil is a vision-based School of Engineering, Indian Institute of Technology Mandi, India1 handwriting interaction device for Khalifa University of Science and Technology Abu Dhabi UAE non-touch screens. • Teleoperation of a 6-DOF serial • Visual sensor (Camera) is directly Hardware design of Visual Pencil manipulator from a 6-DOF parallel mounted in the device. manipulator in master-slave • The device integrates the functions configuration of the mouse and smart pen, thus • Joint-space to Joint-space and task-space removing the requirement of more to task-space mapping is presented auxiliary equipment. • Teleoperation experiment is performed • This work sheds light on the over internet connection and delays are interaction with non-touch screen Conceptual diagram of manipulating Visual Pencil observed computers.

WP1-2(3) 14:00-14:15 WP1-2(4) 14:15-14:30 Learning from Demonstration: Situation-Adaptive Prediction of Continuous Motion for Lower Limb Lane Change Trajectory Planning for Automated Joints Based on SEMG Signal Highway Driving Yongjie Shi 1, Shigang Wang 1, Jian Li 1, Xueshan Gao 1,2 1Robot Institute of Guangxi University of Science and Technology, Liuzhou, China Yulong Liu, Yahui Liu and Xuewu Ji Liting Sun and Masayoshi Tomizuka 2School of Mechatronical Engineering of Beijing Institute of Technology, Beijing, China State Key Laboratory of Automotive Safety and Department of Mechanical Energy, Tsinghua University, Beijing, 100084, Engineering, University of California, Berkeley, • It a prediction model of lower China CA 94720 USA extremity joint continuous motion • the driving situation- based on the combination of GA-BP Expert lane change Driving situation Lane change neural network and limited adaptive lane change demonstrated assessment: lane change trajectory planning utility 233214563( ,  ,...,  ) trajectories in different 12 n cost function amplitude filtering is proposed in trajectory planning driving situations database IRL: learning weighting η , 풞 , η , 풞 , … , 1 1 2 2 ℰ ,ξ , ℰ ,ξ , … , (ℰ ,ξ ) factor 123333333(12 ,  ,..., n ) Offline training Offline 1 1 2 2 n n (η , 풞 ) approach by learning from 푛 푛 this paper. Online optimization expert demonstrations is Driving situation Situation-dependent cost Online lane change • Optimized BP neural network evaluation 123ˆ function selection 123풞 Trajectory optimization proposed. parameters using genetic algorithm • the corresponding situation- Proposed trajectory planning framework Data collection dependent cost function can be synthesized online.

WP1-2(5) 14:30-14:45 WP1-2(6) 14:45-15:00 Design and Experimental Verification on Analysis of Uneven Wear in Driving Mechanism Characteristics of Local Electro-Hydraulic of Flapping-Wing UAV Generation System Jieyu Jia, Weimin Cui, Yugang Zhang, Linjie Shen and Han Wu School of Aeronautics, Northwestern Polytechnical University Yongling Fu, Zhenyu Gou, Mingkang Wang, Deiming Zhu and Ziwang Lin Xi’an, China School of Mechanical Engineering and Automation, Beihang University Beijing, China • Establish the multi-body • Modeling process is following the ‘electric- dynamic model with clearance mechanical-hydraulic’ energy sequences and hinge in LMS Virtual Lab simulative validation was carried based on AMESim platform. • Discretize and parameterize the uneven wear area • A prototype of designed LEHGS was processed and tested. • Simulating the contact stress curves • The LEHGS is able to maintain the pressure with 28MPa and the flow with 14.2L/min • Summarize the uneven wear simultaneously. law of the bushing The LEHGS prototype Flapping-wing driving mechanism • The total efficiency of the LEHGS exceeds 70%.

IEEE ICMA 2020 Conference Digest WP1-3: Industrial, Manufacturing Process and Automation (I)

Session Chairs: Liwen Cao, Heilongjiang University Jian Li, Guangxi University of Science and Technology Online Conference Room 3, UTC+8(Beijing Time): 13:30-15:00, Wednesday, 14 October 2020

WP1-3(1) 13:30-13:45 WP1-3(2) 13:45-14:00 Technology of Automatic Generation and Multi-Level Opportunistic Maintenance Strategy Update of Assembly Dimension Chain for Optimization of Offshore Wind Turbine Based on Assembly Process Rolling Horizon Approach Renchao Zhang, Xiaojun Liu and Yang Yi Hanrong Luo, Chun Su School of Mechanical Engineering., Southeast University School of Mechanical Engineering., Southeast University Nanjing, Jiangsu Province, China Initialize and Determine Parameters Nanjing, Jiangsu Province, China ,,,,,,,,, CPR C C C p p   T Left_bracket Axis Right_bracket Left_bracket Axis Assembly_process_model k k k k0 1 R+ 1 1 R p • A multi-level assembly accuracy Set the Number of Age Groups between Two

Calculate the Original Mean Time to Failure T L and Cumulative Time T C information model is established • A multi-level preventive opportunity i i，0 maintenance model is established. j=0, tj=0 • Assembly information units are j=j+1 Assembly_process_model T

C each maintenance window, multiple levels For Failed Components, update A ij, and T ij , accuracy information model Dimension Chain Dimension Chain Assembly_process_model of maintenance are carried out for the For Running Components, Calculate the Effective Age, Get the Interval of Age Threshold Assembly_process_model A= A + t − t • The automatic generation of i, j i , j−− 1 j j 1

components in different effective age groups Calculate the Maintenance Cost of Each C Assembly relationship transfer Maintenance Cycle j Right_bracket N • Compared with traditional preventive j>10? diagram is realized. Y Calculate Annual Maintenance Cost c • The optimal assembly dimension maintenance methods, the proposed optimization model can significantly reduce chain path is obtained by traversing Generation and Update of multi-level opportunistic the transfer graph. the cost rate. maintenance strategy dimension chain

WP1-3(3) 14:00-14:15 WP1-3(4) 14:15-14:30 Open CNC System Design for Multi-axis Laser Manufacturing Machine Tools Application of Improved Rapidly-exploring Random Zhenshuo Yin, Qiang Liu, Pengpeng Sun, Chenxin Zang and Jian Wang Trees (RRT) algorithm for Obstacle Avoidance of School of Mechanical Engineering and Automation ，Beihang University Snake-like Manipulator Beijing, China Zhen Wang, Jian Chang, Bin Li, Cong Wang and Chun Liu • The Architecture of CNC System. State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Shenyang, China • HMI and Code Translator. • Autonomous obstacle avoidance in • Global Feed rate Planning Module unstructured environment for snake- and Laser Power Optimization like manipulator. Module. • Improved Rapidly-exploring • Interpolator Module and Coordinate Random Trees (RRT) algorithm Transformation Module. • Kinematic model of snake-like • System Instance and Machining Test. BH-LaserCNC320 Machine Tool manipulator and obstacle. • Collision detection and path optimization The snake-like manipulator.

WP1-3(5) 14:30-14:45 WP1-3(6) 14:45-15:00

Survey of Robot Joint Resonance Frequency Satellite Assembly Technology by Robot under Visual Guidance Detection Lijian Zhang, Ruiqin HU and Huajun Chen Hong Wu, Kairan Fan Beijing Institute of Spacecraft Environment Engineering School of Electronic Engineering, Heilongjiang University Beijing, China Harbin, China • Apply vision-guided robots to • Robot joint system modeling satellite assembly. • Resonance mechanism of double • Infrared cameras and "inflexible Camera frame" had been used within the Probe tool inertia joint servo system Robot vision system to degree the goal • Resonance frequency detection pose. method • A method for calculating the Mounting • Key issues of resonance frequency surface pose transformation matrix and detection technology of joint system Target ball Component goal pose of the robotic is given. • Development trend of resonance • The method can meet the detection technology in joint system Manipulator joint model Robot assembly system engineering implementation necessities for satellite assembly.

IEEE ICMA 2020 Conference Digest WP1-4: Vision System, Robotic Vision (I)

Session Chairs: Keigo Watanabe, Okayama University Hsien-I Lin, National Taipei University of Technology Online Conference Room 4, UTC+8(Beijing Time): 13:30-15:00, Wednesday, 14 October 2020

WP1-4(1) 13:30-13:45 WP1-4(2) 13:45-14:00 Inspection and Compensation of Spindle Experimental Eye-in-hand Calibration for Thermal Extension Based on Machine Vision Industrial Mobile Manipulators on ICMA 2020 Longjia Zhang, Xi Zhang, Xiangwei Liu, and Xiezi Guo Zhengxue Zhou, Leihui Li, Xuping Zhang School of Mechatronic Engineering and Automation, Shanghai University Department of Engineering, Aarhus University Shanghai, China Aarhus, Denmark • Studied the application of machine • The mobile manipulator vision in the measurement and compensation of spindle thermal system combines the elongation. mobility and manipulation. • Designed a measurement system • High speed eye-in-hand which consists of image acquisition calibration method. unit and control calculation unit. • High detection rate (94%) • Proposed a method of measuring • and low manipulation error and compensating the thermal (0.89mm) based on the Region elongation of the spindle which can The image acquisition unit growing algorithm and Tsai detect and compensate the thermal calibration method. elongation of the spindle. The Mobile Manipulator

WP1-4(3) 14:00-14:15 WP1-4(4) 14:15-14:30

Towards Robust Auto-calibration for Head- Object Detection Technology of Substation mounted Gaze Tracking Systems Patrol Robot Based on mYOLO Algorithms Meng Liu1, You Fu Li1 and Hai Liu2 1Department of Mechanical Engineering, City University of Hong Kong, Hong Kong Yonggang Zhang, Wenqiang Li, Feng Shen 2National Engineering Research Center for E-Learning, Central China Normal University, China Southern Power Grid Company Limited Harbin Institute of Technology Wuhan, Hubei Province, China Kunming, China Harbin, China

• Adopt an RGBD sensor as Input Collecting calibration data Auto-calibration 3D gaze estimation • Combines multiple groups of the scene camera to capture 3D target data camera image data to realize the fast 3D data of the environment. Scene images Selected scene images Saliency maps recognition and accurate positioning

• Apply the BoW algorithm Extrinsic of the target in the substation. parameters for robust auto-calibration. Depth images Aligned depth data • Provides technical support for the

Gaze vector • Combine the point cloud Pupil Eye vector data smooth operation of substation Eyeball center and rotated gaze vectors to Eye images inspection robot. Eye model Eye camera determine the 3D point of regard. 3D gaze estimation with auto-calibration The Collaborative Positioning System

WP1-4(5) 14:30-14:45 WP1-4(6) 14:45-15:00 2-D Deep Learning Model on 3-D Image Estimating a Driver’s Gaze Point by a Remote Segmentation Spherical Camera Chien-Chia Lee and Hsien-I Lin Zihao Zhao, Shigang Li and Takahiro Kosaki Graduate Institute of Automation Technology, National Taipei University of Technology Mechatronics Research Lab., Hiroshima City University Taipei, Taiwan Hiroshima, Japan • A new segmentation pipeline for • Detecting a driver’s gaze point by a partial view point-cloud based on a single remote spherical camera. 2-D deep-learning model. • Gaze point estimation based on the • Propose an auto-labeling method for geometric model using OpenFace. 3D information. • Use a neural network to estimate the • Compared fairly with PointNet in gaze point in the spherical image. processing time and accuracy. The captured image • The FPS of our method is 20 to 31 times faster than Point-Net.

Predicted point-cloud in different resolutions

IEEE ICMA 2020 Conference Digest WP1-5: Space Robotics and Mechatronics

Session Chairs: Hideyuki Hirata, Kagawa University Wei Zhou, Beijing Institute of Technology Online Conference Room 5, UTC+8(Beijing Time): 13:30-15:00, Wednesday, 14 October 2020

WP1-5(1) 13:30-13:45 WP1-5(2) 13:45-14:00 Study on Control System for Joint-Assisted Flexible Catch Claw-type Space Debris Capture Exoskeleton Under Deep Squat-Rise Process in Mechanism Based on Constant Torque Spring Active Spacesuit Deployment Zhaoyang Li, Yuehong Dai, Peng Tang, and Jiejun Hu Kui Sun, Kailun Chen School of aeronautics and astronautics, University of electronic science and technology of China State Key Laboratory of Robotics and System, Harbin Institute of Technology Chengdu, China Harbin, Heilongjiang Province, China • The resistance torque-angle relation • Space debris cleaning technology can not only curve of joint motion under deep squat- save scarce orbit resources, but also protect rise process with load is obtained satellites and space stations in orbit. • The dynamic equations of three degrees • A design scheme of flexible catch claw-type of freedom model under deep squat-rise space debris capture mechanism based on motion is established constant torque spring deployment is proposed. • An adaptive robust switching learning is • It is mainly composed of a multi-finger flexible presented. catching claw, a transmission device, and a The new flexible catching mechanism • A concept of active spacesuit based on locking device. joint-assisted exoskeleton technology is • It has the advantages of light weight, high presented The prototype of joint-assisted exoskeleton capture reliability and high adaptability.

WP1-5(3) 14:00-14:15 WP1-5(4) 14:15-14:30 A Novel Non-cooperative Target Capture Tool for Development of a Landing Leg with Active Defunct Satellite Buffering and Anchoring Functions Applied to Yongjun Sun*, Yiwei Liu, Minghe Jin, Rongqiang Liu, Hong Liu State Key Laboratory of Robotics and System, Harbin Institute of Technology the Small Body Landing Mechanism Harbin, China Zhijun Zhao, Tao Xiao, Zixin Tang, Xiangyu Gao, Xin Liu, Wenming Zhang, Bin Liu • A novel LAR capture tool is Beijing Key Laboratory of Intelligent Space Robotic Systems Technology and Applications, designed for defunct satellite. Beijing, China • Design a brake for the braking of • Introduction. Bracket Coupling Link capture tool, provide sufficient • Constraints of the small body Gear box Leg stiffness to ensure the fine operation Break landing legs design. Encoder of subsequent maintenance tasks. Torque sensor • Design of the landing leg. Motor/Reducer • Definition coordinate system of • Landing performance test. LAR and capture tool , analysis the • Ultrasonic drill test. Anchoring mechanism kinematics. Foot Anchoring rod • The prototype of the LAR capture • Conclusion. The Capture tool Small body Landing leg tool was developed and tested.

WP1-5(5) 14:30-14:45 WP1-5(6) 14:45-15:00 Design and Implementation of Ultra-Low Speed The Design and Implement for Rotating Optical Path Scanning Control System for telescope and half angle mirror scan Space-Borne Fourier Spectrometer system based on DSP and FPGA Chun Wang, Lanjie Guo, Li Song and Jianbing Kang Li Jing, Zhang Ai, Li Yinlong Department of Control Engineering, Beijing Institute of Space Mechanics & Electricity Beijing Institute of Space Mechanics & Electricity, Beijing, China Beijing, China • An ultra-low speed optical path scanning control system with the • The feature of the system of Rotating angular velocity of 0.1311°/s is designed and implemented for a telescope and half angle mirror space-borne Fourier spectrometer • The design of system architecture system. The experimental results is • The design of hardware and software obtained from the prototype and the velocity stability reaches 99% above, • The experiments and the results which meets the requirement of • Conclusion spectrometer system for high signal- to-noise ratio spectral detection Rotating telescope assembly(RTA) performance. Optical Path Scanning & half angle mirror(HAM) Control System

IEEE ICMA 2020 Conference Digest WP1-6: Biomimetic Underwater Robots

Session Chairs: Yili Fu, Harbin Engineering University Xihuan Hou, Beijing Institute of Technology Online Conference Room 6, UTC+8(Beijing Time): 13:30-15:00, Wednesday, 14 October 2020

WP1-6(1) 13:30-13:45 WP1-6(2) 13:45-14:00 Effect of Straight-wing Pectoral Fin Motion on A Multi-Binocular Camera-based Localization Propulsive Performance of Robotic Dolphin Method for Amphibious Spherical Robots Yangyang Zhu, Xufeng Zhou, Sinan Liu, Yunde Shi and Dan Xia Mugen Zhou, Shuxiang Guo, Liwei Shi, Huiming Xing, Xihuan Hou, He Yin, Debin Xia , Zan Li School of Mechanical Engineering , Southeast University Key Laboratory of Convergence Medical Engineering System and Healthcare, Nanjing, China the Ministry of Industry and Information Technology, Beijing Institute of Technology • Establish the straight-wing pectoral fin • To address position drifts and short positioning dolphin model range in underwater environments, a novel multi- • Evaluate the thrust generated by the binocular camera based localization method with straight-wing pectoral fin through the y x panoramic view for amphibious spherical robots blade element theory z is proposed. • Use fluid calculation software to • A multi-binocular camera system simulate the movement of the robotic dolphin Straight-wing pectoral fin makes the robot observe targets in robotic dolphin model • Analyze the thrust and combined force a nearly 360-degree visual range without generated by the left and right cycloidal rotation movement. propellers in each movement mode • The proposed localization method The cooperative location diagram reveals high accuracy and robustness.

WP1-6(3) 14:00-14:15 WP1-6(4) 14:15-14:30 Design and Analysis for the Axe Foot of Bio- A Modiﬁed Line-of-Sight Method for Path inspired Anchoring Robot Tracking Applied to Robotic Fish Lei Liu1, Linsen Xu2*, Jinfu Liu1, Jiajun Xu1, Gaoxin Cheng1 Sheng Du1,2, Junzhi Yu1,3, and Zhengxing Wu1,2 Shouqi Chen1 ，Hong Xu1，Jia Shi1 1RoboticsState Key Lab Management and Control for Complex Systems, Institute of 1. University of Science and Technology of China, Hefei 230026, China Automation, CAS, Beijing 100190, China 2. Anhui Province Key Laboratory of Biomimetic Sensing and Advanced Robot Technology, 2University of Chinese Academy of Sciences, Beijing 100049, Hefei, Anhui Province, China 3Dept. Mech. Eng. Sci. BIC-ESAT, College of Engineering, Peking University, Beijing 100871, • A novel anchoring robot imitating the razor China clam is proposed. • an improved line-of-sight method is

• This bio-inspired robot can autonomously proposed to navigate a two-joint-actuated Caudal fin Motor controller Rigid body dig holes, anchor, and detach to improve the tuna-like robotic fish. efficiency of anchoring. • The reference head angle is designed with cross-track error, which substantially • The rationality of the screw-cone drill design simplifies the control of linear path tracking. Transmission Motors Pectoral fin Batteries is proved by mechanics analysis. • The obtained simulation results show that mechanism • The capability of obstacle avoidance is the proposed method is able to regulate both The Robotic Fish verified by the bending experiment of the the cross-track error and the heading error soft part. The axe foot fluctuations to an acceptable range quickly.

WP1-6(5) 14:30-14:45 WP1-6(6) 14:45-15:00 Path Tracking Control of Hybrid-driven Robotic Analysis on Driving Force of Flexible Webbed Fish Based on Deep Reinforcement Learning Wings Applied in Wave Glider Liangping Ma, Zhenjia Yue and Runfeng Zhang Baoqiang Tian, Qianjin Wang, and Wenxian Tang International Education College, Zhengzhou University of Light Industry, Zhengzhou, Marine Equipment and Technology Institute, Jiangsu University of Science and Technology, Henan, China. Zhenjiang, Jiangsu Province, China • This article is based on autonomous • Flexible webbed wings (FWWs) learning and autonomous decision- are significant components of wave making control technology, drawing glider to generate driving force. on human learning and decision- • The calculation model of the The prototype of HRF in sail driven mode. making processes, so that the FWWs driving force is established. aircraft can accumulate past control • The analysis on the driving force of experience in a complex marine FWWs under different bending environment, acquire knowledge, deformation conditions is and constantly improve its own completed through the CFD performance and adaptability to method. achieve path following purpose. Target trajectory tracking based on Flow Field Pressure Cloud Diagram reinforcement learning.

IEEE ICMA 2020 Conference Digest WP2-1: Robot Navigation and Control Algorithm (I)

Session Chairs: Kazuhiro Kosuge, Tohoku University Cheng Yang, Beijing Institute of Technology Online Conference Room 1, UTC+8(Beijing Time): 15:15-16:45, Wednesday, 14 October 2020

WP2-1(1) 15:15-15:30 WP2-1(2) 15:30-15:45 An Enhanced Safe and Reliable Autonomous FSM for Robot Target Search and Retrieval under Semi-constructed Environment Driving Platform using ROS2 Xihan Ma, Honglin Sun, Enwei Xu, Song Cui, Boqun Yin, Mariam Faied Hang Cui, Jiaming Zhang, William R. Norris Department of Electrical and Computer Engineering and Computer Science, University of Detroit Mercy, Detroit, MI, US Autonomous and Unmanned Vehicle Systems Laboratory, University of Illinois at Urbana-Champaign, Champaign, US • Developed a multi-tasking robot for object searching and retrieval mission. • Safety-critical Control Application • Adopted finite state machine for • High-low level Controller Framework robot behavior transition • Real-time DBW Kit • The system was deployed under • Robot Operating System 2 ROS environment, programmed using MATLAB • Simulink Controller Modeling • Tested our algorithm on actual hardware in our customized field Jackal UGV configuration The Illi Racecar

WP2-1(3) 15:45-16:00 WP2-1(4) 16:00-16:15 Situation Deduction and Emergency Decision A UWB 3D localization algorithm based on Making Algorithm for Urban Fire based on residual weighting Modified Cellular Automata Method Yonggang Zhang, Qi Liu, Feng Shen, Wenqiang Li China Southern Power Grid Company Limited Harbin Institute of Technology Zijing An, Yingqiu Xu Yingzi Tan Kunming, China Harbin, China School of Mechanical Engineering School of Mechanical Engineering Southeast University, Nanjing, China • An improved algorithm based on Chan • Establish situation deduction method model algorithm to improve the accuracy of three- and analyze the spread of urban fire without dimensional (3D) positioning. any intervention. • The UWB locates the base station group, • Establish the plan optimization method uses the three-dimensional Chan algorithm model and analyze the impact of emergency to estimate the tag coordinates for each rescue activities in different plans. group of base stations, and use the error by UWB using the residual value weight in method. • Establish the dynamic path planning model Positioning System based on deduction data to help rescuers keep away from the dangerous areas. The Fire Deduction Processes under Emergency Plans

WP2-1(5) 16:15-16:30 WP2-1(6) 16:30-16:45 Multi-AUV Terminal Guidance Method Based On Safe Reinforcement Learning with Policy-Guided Underwater Visual Positioning Planning for Autonomous Driving Jingxiang Feng, Yao Yao, Hao Wang, Haoqiang Jin Jikun Rong, Nan Luan Jiangsu Automation Research Institute School of Mechanical Engineering, Shanghai Jiao Tong University Lianyungang, China Shanghai, China

• Use optical beacon array and the AR • Use a hierarchical structure to the change of X-axis the change of Y-axis Marker as the guiding target. 0 0.2 guarantee the safety of

-0.2

m m / / 0

• Determine the pose position and -0.4 reinforcement learning. distance distance -0.2 -0.6 The optical beacon array The optical beacon array The AR Marker The AR Marker orientation of the guiding target The setting value The setting value -0.8 -0.4 • Traffic rules and driving policies are 0 10 20 30 40 0 10 20 30 40 based on the PNP method. time/s time/s the change of Z-axis the change of drift angle described by LTL, and then 4 20 The optical beacon array The optical beacon array The AR Marker 15 The AR Marker

3 The setting value The setting value °

• Design the UUV tracking process m / / constructed a minimum-violation 10 2 angel 5 based on the underwater visual distance 1 problem for low-level policy-guided 0 target. 0 -5 0 10 20 30 40 0 10 20 30 40 planner. time/s time/s • Use three micro AUVs to carry out The information extracted • The low-level planner return the the pool test, which verifies the by the vision algorithm extents of violations as reward. feasibility of the method.

IEEE ICMA 2020 Conference Digest WP2-2: Manipulator Control and Manipulation (I)

Session Chairs: Junjie Zhou, Beijing Institute of Technology Tobias Gold, Friedrich-Alexander-Universitat Erlangen-Nurnberg Online Conference Room 2, UTC+8(Beijing Time): 15:15-16:45, Wednesday, 14 October 2020

WP2-2(1) 15:15-15:30 WP2-2(2) 15:30-15:45

Towards a Generic Manipulation Framework A Biomimetic impedance controller for Robotic for Robots based on Hand Variable Stiffness Grasping Liu, Li Jiang, Shaowei Fan*, and Chongyang Li Model Predictive Interaction Control State Key Laboratory of Robotics and System，Harbin Institute of Technology Harbin, China Tobias Gold, Alexander Lomakin, Tim Goller, Andreas Vö lz, and Knut Graichen • Establishing a linear model between Chair of Automatic Control, Friedrich-Alexander-Universitä t Erlangen-Nü rnberg Erlangen, Germany the endpoint force/stiffness and the human sEMG signal features • Optimization-based control framework • Utilizing the sEMG features of for robotic manipulation tasks corresponding muscles to estimate • Hierarchical decomposition of the endpoint force/stiffness indicators manipulation tasks to a sequence needed for the grasping task of elementary manipulation primitives • Calculating the finger-joint • Generic implementation of these primitives impedance parameters from the by means of a single control structure endpoint force/stiffness and applying it on a biomimetic impedance • Experimental evaluation for a screwing application controller for robotic hand

WP2-2(3) 15:45-16:00 WP2-2(4) 16:00-16:15 Sensorless Collision Detection Method for A novel sampling method for sampling-based Robots with Uncertain Dynamics Based on algorithm Fuzzy Logics used to solve the problem of attitude constraint Yihui Yao, Yichao Shen, Yan Lu, and Chungang Zhuang* Yahao Wang ,Run Yang and Yutao Cheng School of Mechanical Engineering., Shanghai Jiao Tong University CAS Key Laboratory of Mechanical Behavior and Design of Materials, University of Science Shanghai, China and Technology of China, • The dynamic parameters of a 6-DOF industrial robot without extra • Sampling in parallel in the sensors are identified. configuration space and the working space. Greatly • The source of dynamic uncertainty improve the sampling speed. is analyzed. • Treat each sampling as a • A sensorless collision detection differential change, and apply Structure Chart of the Collision method considering the robot constraints on the speed of Detection Method dynamic uncertainty is proposed. pose change. • Fuzzy logics are applied to improve Motion planning under pose constraints the performance.

WP2-2(5) 16:15-16:30 WP2-2(6) 16:30-16:45

An Obstacles Avoidance Algorithm Based on Automated Real Time Image Based Visual Improved Artificial Potential Field Servo Control of Single Cell Surgery Sixi Lu, En Li, Rui Guo Armin Eshaghi and James K Mills School of Artificial Intelligence, University of Chinese Academy of Sciences, No.19(A) Yuquan Department of Mechanical and Industrial Engineering, University of Toronto. Road, Beijing 100049, China Toronto, Canada • Obstacle abvoidance on anipulator • Manipulation of embryonic cells based on improved artificial requires precision and repeatability potential field. • Automation of tasks leads to less • The idea of algorithm is to find a errors and higher throughput path by fitting curve with target point and critical oscillating points. • Image-based visual servo (IBVS) controller detects objects of interest • The conclusion proves that the path and controls robotic manipulators is much smoother by the algorithm based on improved artificial • Developed system is capable of The result of improved artificial performing cell surgery in real-time potential field in this paper. Automated Microinjection potential field algorithm.

IEEE ICMA 2020 Conference Digest WP2-3: Industrial, Manufacturing Process and Automation (II)

Session Chairs: Qinxue Pan, Beijing Institute of Technology Zixu Wang, Kagawa University Online Conference Room 3, UTC+8(Beijing Time): 15:15-16:45, Wednesday, 14 October 2020

WP2-3(1) 15:15-15:30 WP2-3(2) 15:30-15:45 Material Supply Optimization for Just in Time Robust Optimization of Flow Shop Scheduling Assembly Line with Uncertain Travel Times with Uncertain Processing Time Zhang Jiahua1,2*, Li Aiping1, Liu Xuemei1 Qi Sun, Jianping Dou and Canran Zhang 1School of Mechanical Engineering, Tongji University 2 Department of Mechatronics, Wuxi School of Mechanical Engineering Southeast University., Vocational Institute of Arts and Technology, Yixing, 214206, China Nanjing, Jiangsu Province, China • The route dependent the uncertainty Start • Establish a scenario-based mixed Generation of the initial parameter θ is introduced and the population integer linear programming (MILP) material supply planning model for Fitness evaluation robust optimization model. Y Termination Best result assembly line considering interval criteria • Add two more constraints to ensure travel time is established by using N Selection operator Stop the validation of the model. robust optimization method. Crossover operator • Solve the small sized cases by • A genetic algorithm is proposed to Mutation operator LINGO and solve the mid-to-large solve the model. Discrete levy flight sized cases by master-apprentice Elite preservation and • The material supply planning of an updating new population evolutionary algorithm (MAE). engine assembly line is studied. The flowchart of the algorithm • Illustrate the advantage of the MAE The Gant Chart against eight cases.

WP2-3(3) 15:45-16:00 WP2-3(4) 16:00-16:15 Scheduling Optimization of Automated Storage Variation Simulation of Multi-station Flexible and Retrieval System Based on Four-way Shuttles Assembly based on Finite Element Method Jiaorong Song, Mengke Yang, and Xiaoguang Zhou Chao Xu, Chen Luo ,Yijun Zhou and Gang Zhang School of Automation, Beijing University of Posts and Telecommunications School of Mechanical Engineering., Southeast University Beijing, China Nanjing, China • This paper introduces the automatic • Use a new simulation method equipment composition of four-way based upon finite element method. shuttle storage and retrieval system and analyzes the operation flow. • Able to simulate variation from external loads, from contact with • The scheduling optimization model mating part, and from springback is established by using the genetic when the part has been released algorithm. from the external forces. • Finally, the simulation results show • Verify the method through real that the scheduling path obtained experiment on assembling a Multi-station Flexible Assembly after model optimization can Top view of the four-way shuttle manufacturing part from three significantly improve the operating storage and retrieval system efficiency of the system. flat sheet metal.

WP2-3(5) 16:15-16:30 WP2-3(6) 16:30-16:45 Cutting Process and Force Distribution of Three A CMOS Capsaicin Concentration Converter Typical BTA Drills for BTA Deep Hole Drill Yuhua Shi1, Peng Shang1, 2, Shibei Wang1, Nieyong Huang1 with Auto-Sensitivity Control Circuits for 1Engineering University of PAP, Xi’an, Shaanxi, China Sensing Scoville Scale Applications 2School of Mechanical Engineering Xi’an Jiaotong University, Xi’an, Shaanxi, China Cheng-Ta Chiang1, Fang-Chi Hsu1, Wen-Jie Lin1, and Jyh-Cheng Chen2 1Department of Electrical Engineering, National Chia Yi University, Chia Yi, Taiwan • The BTA drilling force measurement 2Department of Food Science, National Chia Yi University, Chia Yi, Taiwan experiment platform is built. • Reasonable distribution of cutting teeth and • The proposed chip can be easily the geometric angle of cutting edge directly utilized to sense capsaicin affect the stability of drilling process and concentration in real time. the shape of chips. • The auto-sensitivity control circuits • The three components of drilling force of could automatically adjust the unit cutting tooth of BTA drill presents the sensitivities under different capsaicin concentration intervals. non-uniform distribution which is increasing Block diagram of the proposed chip with the radius increases. Three Typical BTA Drills

IEEE ICMA 2020 Conference Digest WP2-4: Vision System, Robotic Vision (II)

Session Chairs: Haiyuan Wu, Wakayama University Qiang Fu, Tianjin University of Technology Online Conference Room 4, UTC+8(Beijing Time): 15:15-16:45, Wednesday, 14 October 2020

WP2-4(1) 15:15-15:30 WP2-4(2) 15:30-15:45 Inventory Management using Color Images Automatic Docking Recognition and Location Haiyuan Wu, Peng Li, Jiwei Zhang, and Qian Chen Algorithm of Port Oil Loading Arm Based on 3D Graduate School of Systems Engineering, Wakayama University Wakayama, Japan Laser Point Cloud Linglong He, Yipeng Chen, and Jianghai Zhao • We describes a feature descriptor for University of Science and Technology of China Hefei, Anhui Province, China inventory management of goods using color images. The feature integrates • The SICK LMS511 laser three kinds of pixel information: • Voxel lattice filter and statistical 1) color coordinates UV color space filter 2) color cumulative histogram • RANSAC Algorithm and Euclidean 3) sum of the brightness of same color Clustering Algorithm • The similarity between features is • 3D bounding box algorithm and evaluated by the Bhattacharyya PCL distance where the ratio of achromatic Goods Recognition pixels is used as weights. The Point Cloud

WP2-4(3) 15:45-16:00 WP2-4(4) 16:00-16:15 Analysis of Ranging Error of Parallel Binocular Rotation Axis Calibration of a Turntable Using Vision System Adaptive Weighted-distance Shiqi Gao, Xin Chen, Xiangfei Wu, Tonghui Zeng, Xiaomei Xie * Institute of Astronautics & Aeronautics, University of Electronic Science & Technology of China Xiaofo Liu, Yijun Zhou, Chen Luo, and Gang Zhang Chengdu, China School of Mechanical Engineering, Southeast University • Establish the geometric model of Nanjing, Jiangsu Province, China parallel binocular stereo system. • This paper proposes a new method • Derive the transferred ranging error using adaptive weighted distance based on field of view angle and to improve the calibration of visualize its spatial distribution. rotation axis. • Derive the modified ranging error • With the adaptive weights, rotation based on two-dimension uncertainty axis can be fitted with fewer errors. due to pixel discretization and • A robust error evaluating method visualize its spatial distribution. is developed to measure the • Conduct the ranging experiment to accuracy of proposed method. Rotation axis verify the reasonability of modified Model and Verification ranging error. of Modified Ranging Error

WP2-4(5) 16:15-16:30 WP2-4(6) 16:30-16:45 The Development of Origin Measurement Research on Detection Method for the Leakage System for NC Machine using IoT Technology of Underwater Pipeline by YOLOv3 shinichi Imai Xinhua Zhao, Xue Wang, Zeshuai Du Technology and Information Sciences Department of Natural Science Tokyo Gakugei College of Automation，Harbin Engineering University University Harbin, China Tokyo, Japan • This paper proposes a method for • NC machine tools, it is necessary to detecting oil spill point of underwater Sensing system measure the coordinates of the Image processing system pipeline based on YOLOv3 Image Data workpiece and end mill. • Use denoising, brightness

• There is a problem that a collision Raspberry pi wireless enhancement and other algorithms to ＋ Camera Computer occurs in a contact-type Display system preprocess the submarine pipeline

measurement and the machine tool X= ・・・ image Y= ・・・ is damaged. Z= ・・・ • Experimental results show that the Tablet • We propose a new IoT measurement YOLOv3 network can quickly detect system using image processing and System pipeline vulnerabilities with high Pipeline with leakage points verify its effectiveness. accuracy and low missed detection rate.

IEEE ICMA 2020 Conference Digest WP2-5: Elements, Structures, and Mechanisms (I)

Session Chairs: Xiufen Ye, Harbin Engineering University Ruochen An, Kagawa University Online Conference Room 5, UTC+8(Beijing Time): 15:15-16:45, Wednesday, 14 October 2020

WP2-5(1) 15:15-15:30 WP2-5(2) 15:30-15:45 Optimization Design and Performance Analysis Identification of Permanent Magnet Synchronous of a Shift Manipulator for the Robotic Driver Motor Based on Extended Kalman Filter and Kejin Wang, Zhuang Fu, Xin Feng, Yao Wang Gradient Correction State Key Lab of Mechanical System and Vibration, Shanghai Jiao Tong University Shanghai, China Yan Zhang,Yongli Bi* and Shigang Wang School of Mechanical & Electrical Engineering, Heilongjiang University • A new design of 2-DOFs shift Heilongjiang, China manipulator for the robotic driver is • Significance of parameter Identification Result proposed for automobile tests. identification:Controller tuning. Identifica- Name True value Error Unit • An optimization model is established • Parameters to be identified:stator Tion value Resistanc 2.8 2.85 1.8% Ω to optimize the structural dimension resistance, stator inductance, e Inductanc of the shift manipulator. 0.000835 0.000837 0.24% H permanent magnet flux and mo- e • The experiment results show that the The Shift Manipulator ment of inertia Flux 0.333 0.333 0% Wb shift manipulator can implement the • identification method:Extended Inertia 0.0008 0.00082 2.5% Kg.m2 shift action and satisfy the Kalman Filter and Gradient requirements of automotive tests. Correction

WP2-5(3) 15:45-16:00 WP2-5(4) 16:00-16:15 Stiffness Analysis of a 3-DOF Parallel Physics-of-Failure based Model for Industrial Manipulator with Variable Geometry Platforms Robot Reliability Prediction Xiaoyong Wu, Yujin Wang, Zhaowei Xiang, Ran Yan, Ruizhi Shu and Rulong Tan Jiafan Zhang, Wei Song School of Mechanical Engineering, Chongqing University of Technology Research Center China, Robotics and Discrete Automation, ABB, Shanghai, China Chongqing, China • The stiffness model is established • To state a Physics-of-Failure based based on the Jacobian matrix, with reliability prediction model for consideration of the geometry industrial robot variables of the platforms. • Enable the quantitatively prediction • The dimensionally inhomogeneous of the contributions of each primary stiffness matrix is transformed into failure to the overall robot failure two homogeneous submatrices by rate. singular value decomposition. • A case study on pick-and-place application fully demonstrated the • A comparative analysis is carried Stiffness adjustment out, with regard to the proposed proposed method. Industrial robot and its key components stiffness performance indices. reliability prediction

WP2-5(5) 16:15-16:30 WP2-5(6) 16:30-16:45 Frequency Domain Model Selection for Servo Multisensor-Configuration for Improved Systems ensuring Practical Identifiability Identifiability and Observability of Mathias Tantau, Mark Wielitzka, and Tobias Ortmaier, Institute of Mechatronic Systems, Leibniz University Hanover, Garbsen, Germany Electromechanical Motion Systems Lars Perner, Lenze Automation GmbH, Braunschweig, Germany Mathias Tantau, Mark Wielitzka, and Tobias Ortmaier, Institute of Mechatronic Systems, • Automatic model selection for models of servo systems Leibniz University Hanover, Garbsen, Germany Lars Perner, Lenze Automation GmbH, Braunschweig, Germany • Minimization of the Kullback-Leibler distance guaranteeing practical identifiability of model parameters • Observability and identifiability of servo system models • Sensitivity calculation in • Investigation of the benefit of sensors such as accelerometers frequency domain along the structure • Experimental • Observability of demonstration with nonstable models stacker crane and linear • Experimental validation positioning system of the benefit with stacker crane

Class of candidate models Observability improvement due to accelerometers

IEEE ICMA 2020 Conference Digest WP2-6: Biomimetic Measurement and Control in Robotics

Session Chairs: Yong Yu, Kagoshima University Fusaomi Nagata, Sanyo-Onoda City University Online Conference Room 6, UTC+8(Beijing Time): 15:15-16:45, Wednesday, 14 October 2020

WP2-6(1) 15:15-15:30 WP2-6(2) 15:30-15:45 Cause Exploration of Accuracy Deterioration Development of a Horizontally Movable Multi- with Time Progress in Localization Using a Rotor for Suppressing Load Sway Ryusei Kira, Keigo Watanabe, Isaku Nagai, and Shinsuke Kanda Speckle Odometer Graduate School of Natural Science and Technology, Okayama University, Sotaro Nakata, Isaku Nagai, and Keigo Watanabe Okayama, Japan Graduate School of Natural Science and Technology Okayama University • The objective of this research is to develop a specific multi-rotor to • A system to measure the temperature change inside a speckle odometer is suppress load sway. developed. • This multi-rotor can move in any • The temperature changes near the horizontal direction in the air by laser module and the optical sensor combining thrusts generated from are investigated. the eight rotors. • The obtained data can be used for the • This is fixed directly to the hook and temperature compensation. controlled so as to cancel the load- The Speckle Odometer sway when appearing a sway motion. A Horizontally Movable Multi-Rotor

WP2-6(3) 15:45-16:00 WP2-6(4) 16:00-16:15 A Speed Regulating Controller for Virtual Virtual Passive Bipedal walking on the Irregular Passive Walking Ground using Kinetic Energy Tracking Control Ping Xu and Chunquan Xu* Aiqun Zheng Ping Xu and Chunquan Xu* Aiqun Zheng Department of Control Science & Engineering, Shanghai New Tobacco Product Research Department of Control Science & Engineering, Shanghai New Tobacco Product Research Tongji University, Shanghai, China Institute Co., Ltd., Shanghai, China Tongji University, Shanghai, China Institute Co., Ltd., Shanghai, China • Design two kinds of irregular • Regulating virtual passive walking grounds: a sloped ground composed based on the kinetic energy tracking of a series of changing slopes and a control algorithm. stepped ground. • Realizing accelerating or • The reliability of kinetic energy decelerating to a specified speed, tracking control algorithm is verified and starting a movement from an through numerical experiments. initial resting configuration. • For the stepped ground, virtual • Simulations were carried out and passive walkers can walk stably on the results verified the effectiveness roads with a roughness of about Sloped ground and of the control law. biped robot 0.53% of their leg length. Stepped ground

WP2-6(5) 16:15-16:30 WP2-6(6) 16:30-16:45 Pick and Place Robot Using Visual Feedback Control and Transfer Learning-Based CNN Research on Active Disturbance Rejection Fusaomi Nagata1), Kohei Miki1), Kento Nakashima1), Akimasa Otsuka1), Control of Underwater Manipulator Based on Kazushi Yoshida1), Keigo Watanabe2) and M.K. Habib3) 1)Sanyo-Onoda City University, 2)Okayama University, 3)American University in Cairo Terminal Sliding Mode • CNN&SVM design tool is developed. Jialin He, Yanhui Wei, Yanfeng Zhao, Zhi Zheng, and Jingheng Fu • A transfer learning-based CNN is Department of Automation, Harbin Engineering University Harbin,China designed using the tool to recognize • It uses saturation function instead the orientations of workpieces. of sign function to make control • A pick and place robot is introduced signal more stable. while implementing a visual feedback • This method is insensitive to flow controller and the transfer learning- disturbance, and has strong based CNN to deal with different robustness. shapes of workpieces. • The reliability, stability and • The visual feedback control enables the robot to approach to the position robustness of the proposed control method are verified by comparing Six Dgrees of Freedom just above target workpieces without Underwater Manipulator conducting the robot’s calibration. Pick and place robot with a CNN the three control methods.

IEEE ICMA 2020 Conference Digest WP3-1: Robot Navigation and Control Algorithm (II)

Session Chairs: Shuxiang Guo, Kagawa University Huiming Xing, Beijing Institute of Technology Online Conference Room 1, UTC+8(Beijing Time): 17:00-18:30, Wednesday, 14 October 2020

WP3-1(1) 17:00-17:15 WP3-1(2) 17:15-17:30 A Collision-Free Person-Following Approach Research on Unmanned Motion Planning Based on Path Planning Algorithm for Urban Roads Based on Multiple Lei Pang1,2, Zhiqiang Cao1,2, Junzhi Yu1,3, Weimin Zhang4, Xuechao Chen4 1State Key Lab Management and Control for Complex Systems, Institute of Automation, CAS, Beijing 100190, China 2School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing 100049, China Condition Screening 3Dept. Mech. Eng. Sci., BIC-ESAT, College of Engineering, Peking University, Beijing 100871, China 4Beijing Advanced Innovation Center for Intelligent Robots and Systems, Beijing Institute of Technology, Beijing 100081, China Yizhuo Liu, Liqiang Liu and Xiaohang Yu • A person-following approach based on path planning, which can provide a collision-free Harbin Engineering University and Heilongjiang Experimental High School path and control commands for a mobile robot Harbin, China • Location of the target person and information of obstacles are obtained by a multi-sensor approach (Camera + 3D LiDAR). • Conditionally filter the trajectories, • A* is used to plan global path from current position of the robot to the target person calculate the cost function value and • TEB is employed to plan local optimized path and output control command add to the alternative trajectories set and sort. • Improve the efficiency of existing algorithms. • Use collision detection screening to improve algorithm security. Vehicle Motion Simulation

The schematic diagram of the proposed person-following approach

WP3-1(3) 17:30-17:45 WP3-1(4) 17:45-18:00 Efficent Lane Change Path Planning based on A New Model of Gaps Patching for Side-Scan Quintic spline for Autonomous Vehicles Sonar Images Based on Heading Optimization Zhiyuan Li, Huawei Liang, Pan Zhao, Shaobo Wang, and Hui Zhu University of Science and Technology of China Strategy Peng Li, Yanru Zhang Hefei, China Harbin University of Commerce, Beijing Aerospace Automatic Harbin, China Control Institute，Beijing，China • Lane Change path planning method • Firstly, the heading angles were learned for Autonomous Vehicles (AVs) to decrease the perturbations. driving in structured environments. • A heading optimized model is designed • It builds upon existing method using to patch the gaps in sonar images. quintic spline with second-order • The sonar image was mosaiced with the geometric (G2-) continuity in a Frenet assistance of latitude and longitude coordinate system. information. • This method have the ability to • The proposed method provides better complete complex movements, which feature structural integrity, reduces the can take into account both the gaps in the image and preserves the object boundaries and features. sonar image after gaps curvature and heading angle. Lane Change Path Planning patching

WP3-1(5) 18:00-18:15 WP3-1(6) 18:15-18:30 Modified SDREF for Ocean Observation Research On Compound Control Method Of Satellites Autonomous Navigation Satellite Optical Axis Based On PQ Method Ai Zhang, Zhe Lin NanXing Yan Zhe Lin and YaNing Liu Beijing Institute of Space Mechanics & Electricity, China Beijing Institute of Space Mechanics & Electricity [email protected] Beijing, China • This paper is aiming at the optical • Use SDREF algorithm to estimate the axis of satellite compound control satellites navigation system. systems,presenting a easy • Give the proof that the calculation compensator design method based amount of SDREF is significantly less on the PQ method,which based on than that of EKF when certain conditions the frequency dividing point of are met. satellite control loop and fast • Propose an improved SDREF algorithm steering mirror loop and compound according to the observability of the them effectively. The pointing system Relative location vector precision,track velocity and between two satellites • Numerical simulation stabilization precision of optical axis The satellite optical axis are improved. compound control systems

IEEE ICMA 2020 Conference Digest WP3-2: Manipulator Control and Manipulation (II)

Session Chairs: Yi Liu, Kagawa University Yan Zhao, Beijing Institute of Technology Online Conference Room 2, UTC+8(Beijing Time): 17:00-18:30, Wednesday, 14 October 2020

WP3-2(1) 17:00-17:15 WP3-2(2) 17:15-17:30 Research on Visual Navigation Manipulator for Feasibility Study on Cloud Communication GIS Ultrasonic Partial Discharge Detection Operation for an Interventional Surgery Robot Run Yang, Yahao Wang, Meng Tao and Erbao Dong Taiyun Zhu, Zhong Chen and Wei Yang Yangming Guo, Shuxiang Guo, Cheng Yang Department of Precision Machinery and Precision State Grid Anhui Electric Power Research Instrumentation Institute Key Laboratory of Convergence Medical Engineering System and Healthcare Technology, University of Science and Technology of China Hefei, China School of Automation, Beijing Institute of Technology Hefei, China Beijing, China • The GIS ultrasonic partial discharge detection robot has been designed • A marker and its corresponding detection algorithm is proposed.

• Robot obtain the target position • This paper proposed a novel cloud communication operation for an interventional by binocular stereo camera , and surgery robot. guide manipulator to complete • The cloud server was used to convert the information between the master controller ultrasonic signal acquisition. and the slave operator. Visual navigation manipulator • Experimental results show that the cloud communication system can meet the feasibility requirements of remote vascular interventional surgery

WP3-2(3) 17:30-17:45 WP3-2(4) 17:45-18:00 A Compliant Control Method and Its Ground SCARA Robots Developed with Modular Method Verification Experiment of A Space Robot Daye Chen, Zhaoheng Zeng, Yisheng Guan*, Haifei Zhu, and Tao Zhang Xiaojun Zhu, Kangkang Dong, Hanxu Sun, Bin Liang, Houde Liu Biomimetic and Intelligent Robotics Lab (BIRL), Guangdong University of Technology School of Automation, Beijing University of Posts and Telecommunications, Beijing, China Guangzhou, Guangdong, China • The compliant control algorithms • Summarize the advantages and proposed in this paper improve the ability principles of modular design of robots to comply with the external force method. quickly and stably. • Develop I-typed, T-typed and RP- • This method converts the external force typed joint modules, as the basic into robotic tool velocities to enhance the joint modules. stability and release the impact load of the robot during capturing processes. • Expand the principles of modular design to the traditional SCARA, • The method is verified on the ground developing a M-SCARA system. experimental system and the experimental Configuration of • Conduct an experiment to verify The modular SCARA results show that the proposed method can the experimental platform simulate the weightlessness environment the developed M-SCARA system. and has a good compliance effect.

WP3-2(5) 18:00-18:15 WP3-2(6) 18:15-18:30 Vision-guided Dynamic Object Grasping of Design and Development Of A Flexible Robotic Robotic Manipulators Operative Microscope For Neurosurgical Jie Huang1, Fei Zhang2, Xiangyu Dong1, Ruijin Yang1, Jia Xie3, and Weiwei Shang2 Applications 1. Maintenance Branch of State Grid Anhui Electric Power Company Limited, Hefei, China Arvind Ram Sankar, Ronak Doshi, Aravind Reddy, Kirit Arumalla, Anita Mahadevan, Vikas 2. Department of Automation, University of Science and Technology of China , Hefei, China Vazhayil, Madhav Rao 3. State Grid Anhui Electric Power Company Limited , Hefei, China Surgical and Assistive Robotics Lab, International Institute of Information Technology, • A simple method of object recognition Bangalore, India and localization is introduced. • Enhances the range of view for • A tracking algorithm is designed for neurosurgeons during surgical the moving object by defining the practices attractive vector. • Hollow structure to insert surgical • The grasping experiments are finished tools on an actual 5-DOF robotic • Pulley based system for angular manipulator. Experimental Platform for movement Object Grasping Flexible Robotic • Repositioning Capability Operative Microscope

IEEE ICMA 2020 Conference Digest WP3-3: Industrial, Manufacturing Process and Automation (III)

Session Chairs: Chao Jia, Tianjin University of Technology Yu Song, Tianjin University of Technology Online Conference Room 3, UTC+8(Beijing Time): 17:00-18:30, Wednesday, 14 October 2020

WP3-3(1) 17:00-17:15 WP3-3(2) 17:15-17:30 Research on Simulation and Optimization of Evaluation of Bolt Axial Stress Based on Production Line of Train Wagon Axle Yufei Zhu, Mengke Yang* and Xiaoguang Zhou Nonlinear Ultrasonic Technique School of Automation, Beijing University of Posts and Telecommunications Qinxue Pan, Meile Chang, Shuangyang Li, Xiaoyu Xu, Ruipeng Pan, Yunmiao Zhang Beijing, China School of Mechanical Engineering, Beijing Institute of Technology • Production line with multiple parallel stations and multiple types Beijing, China of processing originals • The relationship model of bolt • Optimization of production line balance length and nonlinear response is • EM-Plant simulation is used to established by controlling analyze the production line variables, and the optimal length status and verify the of bolts is determined. optimization effect. • Within the elastic deformation • Industrial engineering methods range, the nonlinear coefficient is used to analyze the of the bolt increases production line and put Simulation model of processing line monotonously with the variation forward the optimization of the axial stress. Curve of relative nonlinear scheme. coefficient with tensile stress

WP3-3(3) 17:30-17:45 WP3-3(4) 17:45-18:00 Ambiguity-Free Search Method for Precise A Novel Turning Machining Surface Topography Container Positioning Generation Algorithm Wenhao Yang, Yue Liu, and Fanming Liu Xiufeng Liu, Jianhua Ma, Ruolin Wu, Xiang Yan and Yu Huang College of Automation, Harbin Engineering University State Key Lab of Digital Manufacturing Equipment and Technology Harbin, China Huazhong University Science and Technology • The efficiency and precision of Wuhan, China different search steps are provided • Propose a novel 3D surface • The BCAFM method reduces the topography simulation algorithm number of wrong search candidates that directly calculates topography significantly. data according to the actual tool trajectory (including tool’s position The result in the original method • An error threshold method was used and attitude angle) and cutting tool to eliminate the effect of a wrong contour. baseline vector • Take three speciﬁc tool’s deviation • The number of epochs taken by situations into account to increase The Search Result for One using BCAFM to get the final the completeness of the noval accurate position solution is only Epoch of Observation Data algorithm in reality. The result in the novel method half of the traditional AFM

WP3-3(5) 18:00-18:15 WP3-3(6) 18:15-18:30

Study on Freshwater Fish Image Reduced-Order Kalman Filter for Surface Recognition Integrating SPP and Temperature Monitoring of parameter-variant DenseNet Network thermal Systems Hongjun Wang，Yangyang Shi，Youjun Yue，Hui Zhao Tobias Frank, Mark Wielitzka, Tobias Ortmaier Tianjin Key Laboratory for Control Theory and Applications in Complicated Leibniz University Hanover, Institute of Mechatronic Systems System, Tianjin University of Technology, Tianjin , China • Spatial Pyramid Pooling • Dense Neural Network • Large Scale Thermal Modeling • Converged SPP-DenseNet • Model Order Reduction of LPV network Systems • Experimental results and analysis • Parameteridentification • Block diagram of freshwater fish • Surface Temperature image recognition process Estimation using Kalman-Filter Estimated Temperature Distribution Test set accuracy curve

IEEE ICMA 2020 Conference Digest WP3-4: Micro and Nano Systems

Session Chairs: Lingling Zheng, Kagawa University Qingguang Yu, Tsinghua University Online Conference Room 4, UTC+8(Beijing Time): 17:00-18:30, Wednesday, 14 October 2020

WP3-4(1) 17:00-17:15 WP3-4(2) 17:15-17:30 Studies of Electrohydrodynamic Ejection by Performance Evaluation of an Outer Spiral Induced Current Measurement and Optical Microrobot in Pipes in Different Environments Particle Counting Lingling Zheng, Shuxiang Guo and Zixu Wang Tan Fulong1, 2, Yang Baojun1, 2, Zhang Lu1, 2, Tian Jing1, 2, Wang Zhihai1, 2 Graduate School of Engineering, Kagawa University, 1 Faculty of Information, Beijing University of Technology, Beijing 100124, China Takamatsu, Kagawa 761-0396, Japan 2 Key Laboratory of Optoelectronics Technology, Beijing University of Technology, Beijing • Capsule microrobot has the 100124, China potential to replace traditional • Oscillation of induced current endoscopy as a means of diagnosis suggests forming of the Taylor cone, and treatment. and shaking of the meniscus. • A set of three-axes Helmholtz coil • The pulsed output of the optical is used as the magnetically actuated particle counter indicates that a system to generate the uniform droplet is indeed produced. rotating magnetic field. • Combination of the above two • The outer spiral microrobot has measurements may be used for Induced current at the collector Capsule Microrobot System different speeds in the pipe with monitoring the state of the electrode (a), and the output signal from the optical particle counter (b). and without fluids. electrohydrodynamic ejection.

WP3-4(3) 17:30-17:45 WP3-4(4) 17:45-18:00 Design and Simulation of A Novel Three- Depth measurement for the objects with a small Dimensional Multi-Degree-Of-Freedom height using depth-focus-based microscopic Electrothermal Microgripper vision system Guoning Si, Min Ding, Zhuo Zhang, and Xuping Zhang* Yuezong Wang, Xin Zhang, Hao Chen University of Shanghai for Science and Technology The College of Mechanical Engineering and Applied Electronics Technology, Shanghai, China Beijing University of Technology, Beijing, China ⚫ A novel three-dimensional multi- degree-of-freedom electrothermal • Designed the hardware system and microgripper is designed. created the coordinate frames based ⚫ The microgripper is composed of the system. three identical fingers, each having • Proposed the image processing three rectilinear DOFs provided by a method using a weighted evaluation V-shaped electrothermal actuator and principle. a novel 3D U-shaped electrothermal • The results show that the weighted actuator proposed in this work. evaluation algorithm is effective in The process of depth calculation ⚫ Output displacement and force are A novel 3D U-shaped depth calculation. electrothermal actuator analyzed by Finite-element.

WP3-4(5) 18:00-18:15 WP3-4(6) 18:15-18:30 Micro-displacement Amplifying Mechanism of a Real Time Three-Dimensional Image Processing Piezoelectric Single Crystal Actuator and Its with Application to Automated Cell Surgery Motion Characterization Visual Servoing Tianlu Zhang1, Zhangming Du1, Chao Zhou1, Zhiqiang Cao1 , Shuo Wang1and Long Cheng1 Ze Song and James K. Mills 1.Institute of Automation,Chinese Academy of Sciences Laboratory for Nonlinear Systems Control, University of Toronto Beijing, China Toronto, Canada • Real time 3D image processing is The red and black line represented the essential to achieve automated cell input and output displacement variation surgery via visual servoing of the amplifying mechanism. With the • The proposed algorithm utilizes increase of the input voltage, the input SURF, 3D Canny filter and k-means and output micro-displacement of the clustering to estimate the blastomere amplifying mechanism show the 3D centroid synchronous stable linear change • Experimented results from different The result of amplification Blastomere Centroid Estimation during the same frequency. Z-stack images have verified the characteristic by 3D Segmentation of Z-stack algorithm’s performance and Images universal adaptability

IEEE ICMA 2020 Conference Digest WP3-5: Elements, Structures, and Mechanisms (II)

Session Chairs: James K. Mills, University of Toronto Dongdong Bu, Beijing Institute of Technology Online Conference Room 5, UTC+8(Beijing Time): 17:00-18:30, Wednesday, 14 October 2020

WP3-5(1) 17:00-17:15 WP3-5(2) 17:15-17:30 Multi-Criteria Co-Design Optimization of Mechatronic Systems Fariba Rahimi Dept. Machine Design, KTH Royal Institute of Technology Stockholm, Sweden • Three engineering domains as mechanical, control and embedded • This paper aims to offer a way of control implementation are integrated. testing stability when selecting gears, • Non-linear dynamics of a timing belt in order to reduce the weight of a 3- drive with a physical design approach DOF wrist mechanism for a pick- is developed. and-place 3-DOF parallel robot • Sensor quantization model is • This paper suggested a method of integrated inside the control loop. evaluating stability through • Multidisciplinary design optimization comparison of analysis results with Case Study: method is employed to consider theoretical values that are calculated Modeling of Wrist Mechanism Timing belt drive system objectives and constraints from the using formulas. three domains.

WP3-5(3) 17:30-17:45 WP3-5(4) 17:45-18:00 Analysis and Design of a Novel Continuously Research on Rotor Eddy Current Loss Variable Transmission for Robotics Characteristics and Influencing Factors Shouqi Chen1, Linsen Xu2, *, Lei Liu1, Jiajun Xu1, Jia Shi1, Hong Xu1, Jinfu Liu2, Gaoxin Cheng2 Xiliang Yin, Haipeng Geng and Xingcan Liang2 State Key Laboratory for Manufacturing Systems Engineering Xi'an Jiaotong University, Xi’an,China 1 2 University of Science and Technology of China Hefei Institutes of Physical Science, CAS • Derivation and calculation of eddy Hefei, Anhui Province , China Hefei, Anhui Province , China

current loss formula. 0.1 PM_loss sleeve_loss Rotor_loss • A novel continuously variable transmission • Analysis of rotor eddy current loss 0.08 0.06

(CVT) with potential application in robotics. under different working conditions, (W) PM_loss 0.04

0.02

including permanent magnet eddy 0 • The new transmission principle can achieve 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Times (ms)

current loss and protective sleeve 8000 PM_loss a large reduction ratio, and the nested design sleeve_loss 7000 Rotor_loss loss. 6000 theory to achieve a compact structure. 5000

4000 • Exploring the influence of various PM_loss (W) 3000 • The dynamics of the new transmission 2000 1000

0 factors on the eddy current loss of 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 principle. Times (ms) • develop a prototype with hydraulic drive to rotor, such as material conductivity, Rotor Eddy Current Loss adjust the transmission ratio. thickness of protective sleeve and slot width of stator. Virtual Prototype of the CVT

WP3-5(5) 18:00-18:15 WP3-5(6) 18:15-18:30 Stress and Displacement Distribution of High- Load Capacity of Interference Fit Between High- speed Coupling and Shaft for Interference Fit speed Coupling and Shaft Used in Micro Gas Turbine Peng Shang, Shibei Wang, Yuhua Shi, Tongtong Peng and Nieyong Huang Peng Shang, Ziqing Wang, Zhuo Zhou, Yuhua Shi and Nieyong Huang Engineering University of PAP, Xi’an Jiaotong University, Engineering University of PAP, Xi’an Jiaotong University, Xi’an, Shaanxi, China Xi’an, Shaanxi, China • A model of the interference fit is • The load capacity of interference fit established to calculate the stresses between high-speed coupling and and displacements. shaft used in micro gas turbine is determined to two parameters: one • The couple effect of the contact is limit pressure, the other is limit pressure and angular velocity play angular velocity. an important role in analyzing the load capacity of the interference fit . • The coupling’s load capacity should be checked first because the elastic • It is a sound alternative to use the limit pressure and elastic limit thin-walled coupling because it can angular of the coupling are both produce less centrifugal force and Magnitude of Interference lower than those of the shaft under Limit Pressures and Angular unbalance quantity. Velocities of the Connection the same working conditions.

IEEE ICMA 2020 Conference Digest WP3-6: Biomimetic Systems

Session Chairs: Hideyuki Sawada, Waseda Univesity Liang Zheng, Kagawa University Online Conference Room 6, UTC+8(Beijing Time): 17:00-18:30, Wednesday, 14 October 2020

WP3-6(1) 17:00-17:15 WP3-6(2) 17:15-17:30 Design and evalution of an artificial lateral line Development of the Insect-inspired for an amphibious spherical robots Biomimetic Underwater Microrobot for a Yao Hu, Huiming Xing, Liwei Shi, Shuxiang Guo, Xihuan Hou, He Yin, Yu Liu Father-son Robot System Debin Xia, Zan Li, Mugen Zhou Key Laboratory of Convergence Medical Engineering System and Healthcare Technology, the Wenbo Sui, Shuxiang Guo, Liang Zheng, Ruochen An and Awa Tendeng Ministry of Industry and Information Technology, School of Life Science, Beijing Institute of 1Graduate School of Engineering, Kagawa University Technology Takamatsu, Kagawa 761-0396, Japan • we propose a pressure sensor-based • Designed the structure of the robot artificial lateral line for an amphibious platform and the insect-inspired spherical robots.

underwater microrobot. • we carried a experiment to verify the Structure of the amphibious spherical robot • Made the insect-inspired underwater performance of artificial lateral line to microrobot with refillable structure. measure speed of robot. • Verified the functions of microrobot • The result shows the pressure sensor which are automatic tracking and data is highly correlated with the

fixed distance tracking. The Father-son Robot System velocity of the spherical amphibious Data differences between different pressure robot relative to the water flow. sensors at a specific speed.

WP3-6(3) 17:30-17:45 WP3-6(4) 17:45-18:00 Learning Bionic Motions by imitating animals Effect of Pectoral Fin Flapping Motion on Da ZHAO, Sifan SONG, Jionglong SU, Zijian JIANG, Jiaming ZHANG Swimming Performance of Robotic Dolphin The Chinese University of Hong Kong(Shenzhen), Shenzhen, China Xufeng Zhou, Zhihan Li, Yangyang Zhu, Yunde Shi and Dan Xia Xi’an Jiaotong-Liverpool University, Suzhou, China School of Mechanical Engineering, Southeast University Shenzhen Institute of Artificial Intelligence and Robotics for Society, Shenzhen, China Nanjing, China • Establish a dolphin’s body-pectoral- • The Panbot Series are designed for fin model. complementary psychological • Evaluate the force on the pectoral fin rehabilitation therapy through the blade element theory • This paper proposes strategies that • Employe the Computational Fluid enable quadruped robots to imitate Dynamics (CFD) method to obtain the motions extracted from videos. the hydrodynamic performance of • The strategies composed of a basic the robotic dolphin. motion retargeting module and an • Illustrate expected hydrodynamic advanced walking imitation module performance of the elliptical that is combined with SLIP model. Velocity contour of the dolphin The Panbot Series trajectory pectoral fin motion model.

WP3-6(5) 18:00-18:15 WP3-6(6) 18:15-18:30 Trajectory tracking control of hovercraft based Path-Following Control of an Amphibious on Shutting-Backstepping Sliding Model Robotic Fish Using Fuzzy-Linear Model Dan Bai,Mingyu Fu,Yujie Xu,Hanbo Deng Predictive Control Approach Department of Automation, Harbin Engineering University Harbin, China Jie Pan, Jincun Liu, and Junzhi Yu College of Engineering, Peking University • The development and problems of Beijing, China Back-stepping Sliding Model in the research of hovercraft trajectory • A LOS guidance method is utilized to tracking determine the target state. • Using it to solve the differential • A LMPC algorithm is proposed to expansion and input saturation of solve the constrained PF problem. controller • A fuzzy logic controller is used to • The simulation results show that the convert the control variable to rotation problems can effectively avoid angle. Amphibious Robotic differential expansion and input • The effectiveness of the proposed Fish Prototype saturation. The hovercraft algorithm is tested and verified.

Thursday October 15, 2020

Morning Sessions

TA1-1 Control Theory and Application (I) TA1-2 Signal and Image Processing (I) TA1-3 Mobile Robot System (I) TA1-4 Modeling, Simulation Techniques and Methodologies (I) TA1-5 Cognitive Science and Neural Engineering Methods (I) TA1-6 Medical, Biomedical and Rehabilitation Systems (I) TA2-1 Control Theory and Application (II) TA2-2 Signal and Image Processing (II) TA2-3 Mobile Robot System (II) TA2-4 Modeling, Simulation Techniques and Methodologies (II) TA2-5 Cognitive Science and Neural Engineering Methods (II) TA2-6 Medical, Biomedical and Rehabilitation Systems (II)

Thursday October 15, 2020

Afternoon Sessions

TP1-1 Control Theory and Application (III) TP1-2 Signal and Image Processing (III) TP1-3 Mobile Robot System (III) TP1-4 Modeling, Simulation Techniques and Methodologies (III) TP1-5 Cognitive Science and Neural Engineering Methods (III) TP1-6 Medical Robots for Minimal Invasive Surgery (I) TP2-1 Control Theory and Application (IV) TP2-2 Signal and Image Processing (IV) TP2-3 Mobile Robot System (IV) TP2-4 Modeling, Simulation Techniques and Methodologies (IV) TP2-5 Cognitive Science and Neural Engineering Methods (IV) TP2-6 Medical Robots for Minimal Invasive Surgery (II)

IEEE ICMA 2020 Conference Digest TA1-1: Control Theory and Application (I)

Session Chairs: Toshio Fukuda, Beijing Institute of Technology Dingsheng Luo, Peking University Online Conference Room 1, UTC+8(Beijing Time): 9:30-11:00, Thursday, 15 October 2020

TA1-1(1) 9:30-9:45 TA1-1(2) 9:45-10:00 An Effective Model-based Thruster Failure Structural Stabilization for Offshore Platforms Detection Method for Dynamically Positioned Using a Fin Shaped Damper-Mass Ships with a Rule-Based Control Strategy Tongtong Wang, Guoyuan Li, Robert Skulstad, Vilmar Æ sø y and Houxiang Zhang Mina Malek Azari, Jose Victorio Salazar Luces, and Yasuhisa Hirata Department of Ocean Operations and Civil Engineering, Smart Robots Design Lab., Tohoku University Norwegian University of Science and Technology Sendai, Japan Aalesund, Norway • A novel semi-active control system is proposed for pitch stabilization of offshore • Thruster failure will degrade DP platforms in real operational conditions performance and cause immense using a fin shaped damper mass. danger and losses. • A submerged seesaw shaped tuned mass • The model-based thruster failure damper is optimally tuned in passive mode detection and isolation scheme for using Grey Relational Analysis method. DP vessels is presented. • A rule-based control strategy is proposed • Probability analysis is applied to to adjust the stiffness and damping Stabilization system isolate invalid components. coefficient of the stabilization system. The Residuals of Thruster Failure Mode with a fin shaped damper-mass

TA1-1(3) 10:00-10:15 TA1-1(4) 10:15-10:30 Sensitivity-based Model Reduction for In- Analysis and Experimental Research on Process Identification of Industrial Robots Electromagnetic Field Harmonic for High-speed Inverse Dynamics Permanent Magnet Synchronous Motor Daniel Kaczor, Bjö rn Volkmann, Mathias Tantau, Moritz Schappler, and Tobias Ortmaier, Hao Lin, Baisong Yang, Lie Yu, and Dekun Liu Institute of Mechatronic Systems, Leibniz University Hanover, Garbsen, Germany State Key Laboratory for Strength and Vibration of Mechanical Structures, Lars Perner, Lenze Automation GmbH, Braunschweig, Germany Xi'an Jiaotong University, Xi'an, China • Optimal model design and identification of the dynamics • The high-speed permanent magnet of a state-of-the-art industrial robot considering process- synchronous motor (PMSM) related restrictions supported by the air foil bearing is • Identification of the dynamics used for the air compressor. parameters under on-site • Theoretical and simulation analysis conditions was applied to study the • Sensitivity-based assessment of electromagnetic field harmonic. models • The operation experiments of the • Experimental validation with 6-dof high-speed PMSM were driven by CLOOSQRC 350 serial robot Benefit of the resulting model 10 kW at the speed of 90000 rpm. PMSC experiment system

TA1-1(5) 10:30-10:45 TA1-1(6) 10:45-11:00 The Computed Torque Control for an Aerial Reducing Vibration of A Rotating Machine Robot Equipped with Two Coaxial Rotors with Deep Reinforcement Learning Zhang Tao1, Deng Yian1, Hu Fan1, Zhang Xiangqi1, Wei Yaoyao1, Liu Tianlin1, Paul Bolchover2, Having a 2-DOF Tilt Mechanism Yu Tao2, Zheng Guishui2, Chen Rongbing2, Zhang Ping2, Liu Qing2, Luo Dingsheng1 Xiongshi Xu, Keigo Watanabe and Isaku Nagai 1Key Lab of Machine Perception Department of Machine Intelligence Graduate School of Natural Science and Technology, Okayama University Peking University, Beijing 100871, China Okayama, Japan 2Schlumberger • We design a control tool with • A tandem tilted rotor is proposed. several pads, which is used to reduce • Tilt mechanisms with 2-DOF tilt the vibration of a high speed rotating angles are mounted in each rotor. machine; • A simplified model of the tilted dual • We explore the performance of coaxial rotor is used in simulations. different action spaces and • A control allocation problem for a demonstrate the effectiveness of our total of six actuators is solved. The Proposed Aerial Robot approach in reducing vibration of a rotating machine. The control tool

IEEE ICMA 2020 Conference Digest

TA1-2: Signal and Image Processing (I)

Session Chairs: Qin Li, Beijing Institute of Technology He Yin, Beijing Institute of Technology Online Conference Room 2, UTC+8(Beijing Time): 9:30-11:00, Thursday, 15 October 2020

TA1-2(1) 9:30-9:45 TA1-2(2) 9:45-10:00 An Accurate Particle Filter Device Free The Research of Point Cloud Stitching Localization based on Frequency Hopping Method Technique Based on Kinematic Relationship Manyi Wang, Yuang Yang Xiangwei Liu, Xi Zhang*, and Lijing Zhu School of Mechanical Engineering, NanJing University of Science and Technology, NanJing China School of Mechatronic Engineering and Automation, Shanghai University Shanghai, China • Proposing a particle filter method based on frequency • This paper gives a mathematical hopping to improve the model about point cloud stitching accuracy of DFL. relationship and gets a closed-form • Utilizing redundancy of solution. different frequency • Compared with classical techniques measurement information of which use high-precision slide or put each communication link to The RMSE of results of the frequency hopping the mark-points on the object, this improve the localization method and the fixed frequency method proposed technique gains accuracy. considerable flexibility. • Achieving 0.267m accuracy The Reconstruction system with a 38.1% improvement. The average RMSE of the experiment

TA1-2(3) 10:00-10:15 TA1-2(4) 10:15-10:30 RGB-D Image Instance Segmentation Based On Application of Difficult Sample Mining based on Seeded Region Growing Algorithm Cosine Loss in Face Recognition Shi Qiu, Yijun Zhou, Chen Luo and Gang Zhang Enzeng Dong, Yifan Qiao, and Zufeng Zhang School of Mechanical Engineering., Southeast University School of Electrical and Electronic Engineering., Tianjin University of Technology Nanjing, China Tianjin, China

• A novel target detection framework • Facial geometric correction based based on RGB-D images in an on five key points output by unstructured environment. MTCNN. • Use the instance segmentation • This paper improves the original network to produce rough initial loss function and realizes the masks. feature vector emphasizing the • Refine these masks with improved error classification to guide the seeded region growing algorithm. discriminative feature learning. • Results not only maintain high • Use big data samples for feature accuracy, but also contain the extraction model training. Face recognition process instance information of the target. The Segmentation Result

TA1-2(5) 10:30-10:45 TA1-2(6) 10:45-11:00 Analysis of Anchor-Based and Anchor-Free Object Detection Methods Based on Deep Fisheye image Distortion Correction Based on Learning Spherical Perspective Projection Constraint Shujian Liu, Haibo Zhou, Chenming Li, Shuo Wang* Yue Liu, Baofeng Zhang, Na Liu,Hongyan Li,Junchao Zhu Tianjin Key Laboratory for Advanced Mechatronic System Design and Intelligent Control, Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems., Tianjin School of Mechanical Engineering, Tianjin University of Technology University of Technology National Demonstration Center for Experimental Mechanical and Electrical Engineering Tianjin, China Education, Tianjin University of Technology • Anchor-Based and Anchor-Free • A fisheye correction algorithm detectors based on deep learning based on spherical perspective are summarized,whose projection constraint was proposed. performances are compared and • Through the experiment, this analyzed as well. method has a good effect on the • The development of object correction of fisheye image. detection’s key technologies are summarized. • The development trend of object Branch diagram of Spherical coordinate positioning method detection is discussed. Deep learning-Based detectors

IEEE ICMA 2020 Conference Digest TA1-3: Mobile Robot System (I)

Session Chairs: Keigo Watanabe, Okayama University Lanyong Zhang, Harbin Engineering University Online Conference Room 3, UTC+8(Beijing Time): 9:30-11:00, Thursday, 15 October 2020

TA1-3(1) 9:30-9:45 TA1-3(2) 9:45-10:00 Fuzzy Control-based Three-Dimensional Rapid Generation of Challenging Simulation Scenarios for Autonomous Vehicles Based on Adversarial Test Motion Planning of an Amphibious Xiaokun Zheng1,2 Huawei Liang1,3,4*, Biao Yu1,3,4*, Bichun Li 1,3,4 , Shaobo Wang 1,2, Zhilei Chen 1,2 1. Hefei Institutes of Physical Science, Chinese Academy of 3. Anhui Engineering Laboratory for Intelligent Driving Technology Sciences, Heifei, Anhui, China 230008 and Appllication, Hefei, Anhui, China, 230008 4. Innovation Research Institute of Robotics and Intelligent 2. University of Science and Technology of China, Heifei, Anhui, Spherical Robot Manufacturing, Chinese Academy of Sciences, Hefei, Anhui, China China 230026 230008 Liang Zheng, Shuxiang Guo, Yan Piao, Ruochen An, Wenbo Sui [email protected] Corresponding to: [email protected] and [email protected] Graduate School of Engineering, Kagawa University, Takamatsu, Kagawa 761-0396, Japan

• Fuzzy control realizes floating and diving • 3D underwater movement • Modeling and simulation of • A framework of rapid generation • Use quantum genetic algorithm to control algorithms of challenging simulation scenarios solve the optimization problem owing based on adversarial test was proposed. to its high search ability and lower • Underwater experiment Structure of the spherical • An optimization model based on population size required. newly developed indicators, i.e., safety, • A simulation system based on amphibious robot smoothness, sharpness and smartness, Prescan, Carsim and Simulink was was proposed to assess the autonomy built to validate the effectiveness of of autonomous vehicles. the proposed method. Framework of the simulation test system

TA1-3(3) 10:00-10:15 TA1-3(4) 10:15-10:30 Research and Development of Vision Automated Guided Logistics Handling Vehicle Positioning System for Evaporator Tube Plate Path Routing under Multi-Task Scenarios Crawler Haoxuan Kan, Liguo Shuai, and Huiling Chen, Wenzhe Zhang School of Mechanical Engineering , Southeast University, Keqing Wang,Wenpeng Yuan,Lanyong Zhang Nanjing China College of Automation, Nanjing University of Information Science & technology Bingjiang College Nanjing,China College of Automation，Harbin Engineering University Harbin, China • Automated guided vehicles（AGV） Chart Target error routing algorithm under multi-task Positioning Positioning error in Positioning error in • The improved median filtering Serial error in X number Y direction（mm） Z direction（mm） scenario direction（mm) algorithm is designed and the 1 8.43 6.79 -11.31 2 4.63 8.54 -9.59 • Using Q-learning to solve the recognition method based on 3 7.64 9.11 -10.38 "curse of dimensionality" problem. 4 8.41 8.57 -10.96 BP neural network is adopted 5 8.56 7.34 -9.89 6 5.18 7.01 -8.41 7 8.28 7.20 9.89 • Considering the time window, • Combined with Zhang 8 8.74 7.64 10.39 obtaine the optimal routing calibration method and 3D 9 2.52 6.21 -9.52 sequence using DFS. The error between the measured Fig. 1 Intelligent distribution system structure. reconstruction principle, the 3D coordinates of the target pipe target world coordinates and the actual world coordinates hole contour center are obtained.

TA1-3(5) 10:30-10:45 TA1-3(6) 10:45-11:00 Realization of a Bio-inspired Cheetah Robot with On Task-Specific Redundant Actuation of a flexible Spine Spring-Assisted Modular and Chunlei Wang, Chunhui Fan, Ya Yang, Minghao Zhan, Haicun Shao, and Baoping Ma Reconfigurable Robot 21st Research Institute of China Electronics Technology Group Corporation Shanghai, China Christopher Singh, Student Member, IEEE, and Guangjun Liu, Senior Member, IEEE Aerospace Systems and Control Laboratory, Ryerson University • A bio-inspired cheetah robot is Toronto, Canada presented in this study. • Multiple working mode actuation • Introduction of the original intention approach to uncertain manipulation of this paper. • Reduce motor torque for energy • Introduction of the mechanical savings in ordinary task phases design of the robot. • Task-specific innovation aid for • Introduction of the control algorithm critical effort and second-attempts and control system hardware. • Spring modules are reconfigurable Cheetah robot • Introduction of the conclusions of on-line and offline, and comply with this study. MRR structural reconfiguration Spring-Assisted MRR

IEEE ICMA 2020 Conference Digest TA1-4: Modeling, Simulation Techniques and Methodologies (I)

Session Chairs: Junjie Zhou, Beijing Institute of Technology Yili Fu, Harbin Engineering University Online Conference Room 4, UTC+8(Beijing Time): 9:30-11:00, Thursday, 15 October 2020

TA1-4(1) 9:30-9:45 TA1-4(2) 9:45-10:00 Optimization Algorithm of Battlefield Medicinal Materials Forward Kinematics Solution of Inward Bending Distribution Route Planning 3-RRS Parallel Mechanism Based on Ant Colony Wenming Zhou, Lin Zheng, Yaowu Wu, Yunhe Wang and Dekang Cui Joint Logistic College., National Defense University of PLA Algorithm Beijing, China Shiquan Ni, Dapeng Tian, Dan Qiao, Jiangtao Zheng • Consider the different requirements for 140 1.Key Laboratory of Airborne Optical Imaging and Measurement, Changchun Institute of Optics,

120 3 Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, China medicinal supply support in peacetime, 4 100 2.University of Chinese Academy of Sciences, Beijing 100049, China prewar preparations and wartime. 9 18 80 19 5 • Ant colony algorithm is used to solve the forw- 1 Distribution center • the optimization of the distribution 2 16 60 7 8 17 12 ard kinematics solution, avoiding the influence route of medicinal materials with the 20 10 15 40 11 of the initial value on the results. The ant colo- objective of shortest distance , lowest Supplystation ordinate 6 20 14 ny algorithm can find the global optimization cost and shortest time is studied 13 0 and avoid falling into the local optimum. respectively. 0 20 40 60 80 100 120 140 Supply station abscissa • It does not need to solve Jacobian matrix of • Based on ant colony optimization Coordinates plan of a battle mechanism. When the mechanism is in a sing- algorithm, the optimization algorithm rescue medicine support ular position, Newton and Newton dwon-hill for drug delivery under different station method failing, the algorithm can still obtainInward bending 3-RRS constraint conditions in peacetime and the forward kinematics solution of mechanism. parallel mechanism war is designed.

TA1-4(3) 10:00-10:15 TA1-4(4) 10:15-10:30 A Novel Harmonic Analysis Method of Active A Novel Trajectory Predicting Method of Catheter for the Vascular Interventional Surgical Robot Distribution Network with Multiple Distributed Jian Guo1, Yue Sun1，ShuxiangGuo1,2 1 Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Biomedical Robot Laboratory, Tianjin University of Technology, China Generation 2 Department of Intelligent Mechanical Systems Engineering, Faculty of Engineering, Kagawa University, Japan Tianyi Ma，Yanping Du and Yuan Zhang school of mechanical and electrical engineering, Beijing Institute of Graphic Communication Beijing, China • In this paper, a trajectory prediction system for catheter based on • The modelling method of active Kalman filtering algorithm is proposed. distributed network with multiple • In this paper, the actual trajectory of DGs was established. the catheter is fitted into a curve by fitting method. • The multiple DG harmonic • The experimental results show that calculation method was proposed. the trajectory prediction system can accurately predict the next time • The proposed harmonic calculation trajectory of the catheter. And can greatly improve the safety of the method is verified by experiments, operation. and the experimental results prove that the proposed harmonic Distribution network topology The steps of trajectory prediction calculation method has high accuracy.

TA1-4(5) 10:30-10:45 TA1-4(6) 10:45-11:00 Wind Turbine Performance Optimization Based Flameproof permanent magnet inverter motor on Statistical Data design based on Maxwell Hong Yu and Haiyang Zhou Yunnan Power Grid Co., Ltd. Electric Power Research Institute Lei wang;Hao liu;Yuqin zhu;Xinming wang;Sen wang ;Zeyu Huang;Yujia Li;Di Tang Kunming, China Shenyang Research Institute of Coal Science Group Co.LTD,Key Laboratory of Coal Mine Safety Technology,Shenyang Institute of Engineering • Obtain the main factors affecting the Shenyang,China power generation of wind turbines by • According to the design requirements of the using SVM algorithm and similarity permanent magnet motor and the actual situation, index. combined with the requirements of the explosion-proof motor. • The time series ARIMA prediction • The detailed design including the model is combined, and the power electromagnetic and the finite element analysis obtained by the method is weighted of the prototype design was carried out using with the power obtained by the ANSYS Maxwell software to obtain the no-load previous integration method, so that Comparison of Predicted magnetic density of the permanent magnet the results are improved. Power and Actual Power inverter integrated machine and the magnetic Permanent Magnet density of the short-circuit fault. Motor

IEEE ICMA 2020 Conference Digest TA1-5: Cognitive Science and Neural Engineering Methods (I)

Session Chairs: Jian Guo, Tianjin University of Technology Qiong Wu, Suzhou University of Science and Technology Online Conference Room 5, UTC+8(Beijing Time): 9:30-11:00, Thursday, 15 October 2020

TA1-5(1) 9:30-9:45 TA1-5(2) 9:45-10:00

Temporal Factors for the Flanker Category Neural Decoding Based on Active Learning for Effect under Visual Crowding Intracortical Brain-Machine Interfaces Yiyang Yu1, Qiong Wu2 3*, Jiajia Yang3, Satoshi Takahashi3, Weihua Wang, Jian Huang*, Dongrui Wu, Peng Zhang Yoshimichi Ejima3 and Jinglong Wu3 School of Artificial Intelligence and Automation 1 The Graduate School of Natural Science and Technology，Okayama University，Okayama, Japan; 2 School of Huazhong University of Science and Technology (HUST) Education, Suzhou University of Science and Technology, Suzhou, China. 3 Graduate School of Interdisciplinary Wuhan, Hubei, China Science and Engineering in Health Systems, Okayama University，Okayama, Japan • Proposed three greedy sampling • The aim of this study was to investigate active learning method (named the exposure time for the flanker DGSx, DGSy and iDGS) to solve category effect. the problem of label samples in • Participants were asked to identify the iBMIs. numbers in different crowding • One-class-SVM was assumed to conditions. denoise to achieve stable and robust • The experimental data suggest that the decoding peak of the flanking category effect occurs at the time window of 100-150 • Compare with QBC and Uncertainty Comparison of Accuracy in Sampling, the proposed methods ms after the stimulus appears. Classification Problems achieve higher accuracies.

TA1-5(3) 10:00-10:15 TA1-5(4) 10:15-10:30 Research and development of the Influence of EMG Data from Different Days on synchronization device of fMRI scanning cycle Feature Selection Methods Anyuan Zhang and Qi Li * and external stimulus presentation equipment School of Computer Science and Technology, Changchun University of Science and Technology, China Yulong Liu, Jiajia Yang, Qiong Wu, Hongtao Yu, Satoshi Takahashi, Yoshimichi Ejima, Jinglong Wu • The EMG feature we selected in the The Interdisciplinary Science and Engineering in Health Systems Okayama University first day had a poor performance on Okayama, Japan the other day due to the feature • To solve the problem of information space changes. synchronization between two • However, few studies have focused different hosts. The Psychtoolbox Periodic scan transformation waits for on the performance of feature monitor port • The system is composed of signal output signals selection methods for EMG pattern acquisition and conversion output. recognition when EMG data sets • The high dynamic response and very Monitoring Send run from different days. trigger signal command low delay of the system solve the • We discussed the influence of EMG problem of the operation of data from different day on feature synchronous control of the external System structure selection methods to find a robust device between the two hosts. feature selection method.

TA1-5(3) 10:30-10:45 TA1-5(4) 10:45-11:00

Effect of attentional load on audio-visual Robust Audio-visual Enhancement integration: an ERP study elicited by Exogenous Bimodal Cue Yanna Ren1, Zhenhua Zhou1, Junhao Bi1, Junyuan Li1, Tao Wang3, Weiping Yang2 Yanna Ren1, Ying Zhang1, Bixin Yuan1, Jia Yu1, Tao Wang3, Weiping Yang2 Department of Psychology, College of Humanities and Management, Guizhou University of Department of Psychology, College of Humanities and Management, Guizhou Traditional Chinese Medicine, Guiyang, China. University of Traditional Chinese Medicine, Guiyang, China. • The exogenous cue speeds up • A dual task was conducted to the response to the following investigate the influence of target. attentional load on AVI effect. • The audiovisual enhancement • The AVI theta and alpha was greater in bimodal cue power under low-attentional- condition. load condition. • This result suggested that • This result suggested that the exogenous bimodal cue can AVI effect was greatly affect capture more constant and Topographic maps for theta and alpha activity by attentional load. [AV - (A + V)] greater attentional resource. Response enhancement in different conditions

IEEE ICMA 2020 Conference Digest TA1-6: Medical, Biomedical and Rehabilitation Systems (I)

Session Chairs: Kazuhiro Kosuge, Tohoku University Yi Liu, Kagawa University Online Conference Room 6, UTC+8(Beijing Time): 9:30-11:00, Thursday, 15 October 2020

TA1-6(1) 9:30-9:45 TA1-6(2) 9:45-10:00

Study on the Tremor Elimination Strategy for the Vascular Preliminary Design of a Novel Home-based Tele- Interventional Surgical Robot rehabilitation System for Upper Limb Rehabilitation Yi Liu, Shuxiang Guo and Ziyi Yang 1 1 1,2 Jian Guo , Shuai Yang Shuxiang Guo * The Graduate School of Engineering, Kagawa University 1 Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Intelligent Robot Laboratory Kagawa, Japan Tianjin University of Technology ，Tianjin, China 2 Department of Intelligent Mechanical Systems Engineering , Faculty of Engineering Kagawa University , Japan • In the proposed tele-rehabilitation • A tremor elimination strategy was system, there is a master side for the proposed to filter out the tremor therapist and a slave side for the of hands of doctors who use the patient. vascular interventional surgical • A TCP/IP communication is used to robot transmit the data and feedback • Hands' tremor was divided into between the master and the slave two types , the first type is normal side. physiological tremors and the • The interface is shown on a PC for Overview of the Proposed Tele- second type is stress tremor. the therapist to remotely operate the rehabilitation System Results of tremor elimination exoskeleton worn by the patient.

TA1-6(3) 10:00-10:15 TA1-6(4) 10:15-10:30

Comparison of Isometric Force Estimation Design and Analysis of a Wearable Exoskeleton Methods for Upper Limb Elbow Joints Upper Limb Rehabilitation Robot Ziyi Yang, Shuxiang Guo and Yi Liu Department of Intelligent Mechanical Systems Engineering, Graduate School of Engineering, Kagawa University Jian Guo1 , Pengyu Li1 and Shuxiang Guo1,2* Takamatsu，Japan Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Biomedical Robot Laboratory., Tianjin University of Technology Beijing, China • Use electromyograph (EMG) signals from the biceps to estimate the upper • Building 3D models of the robot limb elbow joint force • Mathematical modeling of the robot

• Comparison of estimation performance using the D-H parametric method between the model-free method and the • Simulation analysis of the robot using model-based method ADAMS simulation software Fig. 1. The framework of proposed bilateral training strategy using sEMG-based intention • The model-free method have the better interpretation for real-time bilateral control. • The simulation results show that the prediction performance which can provide acceptable prediction results motion curve of the robot is smooth, with root-mean-square error (RMSE) which verifies the rationality of the The 3D model of the robot error below 2.20 N robot structure.

Time(s) Fig. 2 Comparison results of model-based and model-free method

TA1-6(5) 10:30-10:45 The design and evaluation of complex radial basis functions network adaptive robust variable sliding mode controller for stroke rehabilitation robot Peng Zhang, Junxia Zhang* Tianjin Key Laboratory for Integrated design & Online monitor Center of Light Design and Food Engineering Machinery Equipment., Tianjin University of Science &Technology，Tianjin, China

• A complex radial basis functions network adaptive robust variable sliding mode controller with healthy-dyskinetic side coordination for active stoke lower limb rehabilitation robot was proposed. • Compared to PID, the higher The simulation of the proposed performances of the proposed controller controller were demonstrated by co- simulation.

IEEE ICMA 2020 Conference Digest TA2-1: Control Theory and Application (II)

Session Chairs: Peng Shi, Kagawa University Jian Li, Guangxi University of Science and Technology Online Conference Room 1, UTC+8(Beijing Time): 11:15-12:15, Thursday, 15 October 2020

TA2-1(1) 11:15-11:30 TA2-1(2) 11:30-11:45 Design of Wave Filter for Dynamic Positioning Model Reference Scheduling and Control Co- Ships Based on Double Kalman Filter (DKF) design with Random Delay Shunli Zhao Guoqing Xia and Binyuan Xia 1. Tianjin Key Laboratory for Control Theory & Applications in Complicated System, Tianjin Harbin Engineering University University of Technology, Tianjin, China Harbin, China 2. School of Electrical and Electronic Engineering, Tianjin University of Technology, Tianjin, • The DKF is composed of the first- China stage Kalman filter based on linear • A study of model reference time-varying systems and the scheduling and control co-design second-stage linearized Kalman with random time delay is proposed filter. in the paper. • Wave filter based on DKF can filter • A pole placement method is used to out high-frequency oscillations and design the controller gain for the estimate the speed. ideal dynamics. • the convergence rate of the Wave • The effectivity is demonstrated by a filter based on DKF is faster than Simulated position chart numerical example. Scheduling scheme in NCS that of the Wave filter based on EKF.

TA2-1(3) 11:45-12:00 TA2-1(4) 12:00-12:15 Performance Optimal and Robust Design of an Control Strategy of Grid-connected Wind Power Idle-Speed Controller Considering Physical Inverter Based on Linear Active Disturbance Uncertainties Rejection Control Eduard Popp*, Mathias Tantau*, Mark Wielitzka*, Dennis Giebert** and Tobias Ortmaier* Yulong Chen, Xuesong Zhou and Youjie Ma *Institute of Mechatronic Systems, Leibniz University of Hanover, Garbsen, Germany Key Research Laboratory for Control Theory & Applications in Complicated Systems **IAV Automotive Engineering, Gifhorn, Germany Tianjin University of Technology Tianjin, China

r + u0 1 u y • Methodology for automated design k G(s) • DC bus voltage control strategy of p b - - - 0 of any controller structure grid-side inverter in permanent magnet k b d 0 • Design in compliance with given direct-drive wind power system. 1 + L xˆ s 1 performance requirements and • The ESO-based LADRC control 1 + 1 + + + parametric uncertainties. L2 technology is used to replace the xˆ s + - 2 • Reducing the simulation effort by traditional PI control , which is applied 1 L using the parameter space to the voltage outer loop control in the xˆ s 3 3 Performance Level of double closed loop control, and approach and a sensitivity analysis Requirements performance achieves good control effects. The structure diagram of LADRC • Approach applied on a vehicle and robustness idle-speed controller

IEEE ICMA 2020 Conference Digest TA2-2: Signal and Image Processing (II)

Session Chairs: Yu Song, Tianjin University of Technology Shinichi Imai, Tokyo Gakugei University Online Conference Room 2, UTC+8(Beijing Time): 11:15-12:15, Thursday, 15 October 2020

TA2-2(1) 11:15-11:30 TA2-2(2) 11:30-11:45 Zebrafish Larva Heart Localization Robust Tensor Principal Component Analysis Using a Video Magnification Algorithm Based on F-norm Mingyu Cui, Liying Su, Peng Zhang, 1,2 2 2 Zhuo Zhang Xuping Zhang* Weimin Ge , Jinxiang Li and Xiaofeng Wang Huipeng Zhang, and Hongmiao Wei 1 National Demonstration Center for Experimental Mechanical College of Mechanical Engineering School of Modern Posts, Department of and Electrical Engineering Education and Applied Electronics Technology, Xi'an University of Posts & Engineering, Aarhus 2 Tianjin Key Laboratory for Advanced Mechatronic System Beijing University of Technology Telecommunications University Design and Intelligent Control Tianjin University of Technology, China Beijing, China Xi’an, China Aarhus, Denmark • A new robust tensor NPC 10 20 30 40 50

• Using a video magnification algorithm algorithm is proposed TPCA-L1-G to solve the problem of detecting the based on F-norm. position of zebrafish larvae with weak • The algorithm can TPCA-L1-NG heart signal under the microscope effectively reduce the • By amplifying the dynamic error of image MPCA A reconstructed image information of the larval heart, the reconstruction with sample of the GT dataset signal of the heart can be directly good classification F-TPCA recognized, and finally the position The Whole Process of rate. coordinates can be obtained. Determining the Heart The reconstructed image of GT dataset

TA2-2(3) 11:45-12:00 TA2-2(4) 12:00-12:15 Research on Detection Technology of Various Fruit Disease Spots Based on Mask R-CNN UAV aerial image mosaic algorithm based on HongJun Wang, Qisong Mou, Youjun Yue, Hui Zhao Tianjin University of Technology FAST-Tomasi feature and Delaunay triangulation Tianjin China Hui Zhao, YaXin Du, Hongjun Wang and Youjun Yue • This paper takes apple, orange, Tianjin Key Laboratory for Control Theory and Applications in Complicated System, peach, and pear as the research Tianjin University of Technology Tianjin, China object, and proposes an improved Mask R-CNN target detection • Fast-Tomasi feature extraction. algorithm. • Feature point elimination based • Based on the recognition and on Delaunay triangulation. localization of the fruit picking robot, • Improved SPHP algorithm for the fruit surface spots were detected. image registration. • Improve Mask R-CNN for multi- • Weighted fusion algorithm for scale feature fusion: add a bottom- image fusion. After Delaunay triangulation up horizontal connection path to the The Detection Model eliminates matching points original feature pyramid structure.

IEEE ICMA 2020 Conference Digest TA2-3: Mobile Robot System (II)

Session Chairs: Shuoxin Gu, Chengdu University of Information and Technology Xueshan Gao, Beijing Institute of Technology Online Conference Room 3, UTC+8(Beijing Time): 11:15-12:15, Thursday, 15 October 2020

TA2-3(1) 11:15-11:30 TA2-3(2) 11:30-11:45

Distributed Edge-based Event-driven Consensus Formation Finite-Time Observer-Based Robust Continuous Strategy for Multiple Amphibious Spherical Robots Twisting Control for the Attitude of an Uncertain He Yin,Shuxiang Guo, Liwei Shi, Huiming Xing, Xihuan Hou, Zan Li, Debin Xia, Mugen Zhou School of Life Science,Beijing Institute of Technology Quadrotor UAV Subjected to Disturbances Beijing, China Omar Mechali, Limei Xu, Abdelkader Senouci, Xiaomei Xie, Chen Xin, Abdesslam Mechali • The proposed formation control School of Aeronautics and Astronautics, University of Electronic Science and Technology of strategy merges virtual linkage China, Chengdu, China formation control algorithm, edge- • This paper deals with robust attitude based event-driven consensus control problem of the quadrotor in control algorithm and artificial the presence of model uncertainties potential field algorithm together. and external disturbances. • The proposed formation control • A Disturbance Observer-Based strategy not only realizes the Control method is proposed by Fig. 1 Experimental setup: Quadrotor platform mounted on a 3-DoF spherical cardan joint. formation control of multiple robots, combining an Improved Twisting but also reduces the driving The Amphibious Spherical Robot Control algorithm and a Finite-Time frequency of each robot Observer. • Experimental results are presented. Fig. 2 Attitude tracking (Φ,θ,ψ).

TA2-3(3) 11:45-12:00 TA2-3(4) 12:00-12:15 Visual perception and navigation of security robot An Approach for Multi-Objective Avoidance Using Dynamic Occupancy Grid Map based on deep learning Xiuzhi Li 1 2 , Kangkai Guo 1 2 , Tong Jia 1 2 , Xiangyin Zhang 1 2 Yanjie Liu and Jiao Chen Xinyu Bai 1 Faculty of Information Technology, Beijing University ofTechnology, State Key Laboratory of Robotics and System Sinopec International Business Co., Ltd. Beijing 100124, China Harbin Institute of Technology, Harbin, China Beijing, China 2 Engineering Research Center of Digital Community, Ministry of Education, Beijing 100124, China • Define the environment as a random dynamic system DOGM. • Road regions are detected based on • Obtain the velocity of multiple deep semantic segmentation network. objects with low delay (71ms) under GPU parallel acceleration • Based on the segmented regions, an calculations. edge extraction algorithm is designed • Expand the data structure of layered to extract and fit the road boundaries. costmaps in ROS stack to connect the path planner and dynamic grid • The proposed technique is verified in map. Dynamic Obstacle Avoidance real environment. Outdoor robot navigation framework

IEEE ICMA 2020 Conference Digest TA2-4: Modeling, Simulation Techniques and Methodologies (II)

Session Chairs: Qingguang Yu, Tsinghua University Huiming Xing, Beijing Institute of Technology Online Conference Room 4, UTC+8(Beijing Time): 11:15-12:15, Thursday, 15 October 2020

TA2-4(1) 11:15-11:30 TA2-4(2) 11:30-11:45

HTN Guided Adversarial Planning for RTS Games Parameters Optimization of PID Controller for an Lin Sun, Anshi Zhu, Bo Li and Xiaoshi Fan Artillery Coordinator Based on Multi-Objective Joint Logistics College, National Defense University Beijing, China Genetic Algorithm Junhua Chen, Longmiao Chen, Wei Miao and Guangsong Chen • The challenges of RTS game for AI Max1 Nanjing University of Science and Technology research are discussed. Nanjing, Jiangsu Province, China Instruction

• The proposed model considers the • A multi-objective optimization analysis Angle Max2 Max3 Controller time constraint when generating the method to decrease the mechanical vibration Encoder

subgame tree by HTN planning. of the artillery coordinator is proposed PS Pr U • The opponent strategy is introduced Max4 Max5 Min1 Min2 • The mathematical multidisciplinary Q P P Q during planning as the adversarial • model of the artillery is established 1 1 2 2 player. Min3 Min4 • Realization of multi-objective genetic algorithm for coordinator The Game Tree Example model of the coordinator • Performance of the method demonstrated with some cases.

TA2-4(3) 11:45-12:00 TA2-4(4) 12:00-12:15 Energy Consumption Data Prediction Analysis Integrated Design and Verification Method of based on EEMD-ARMA Model Aircraft and Propulsion System Ji Li, Xudong Jiang, Lei Shao, Hongli Liu, Chao Chen, Gang Wang and Dongdong Du Wang min, Gu wei-qun, Sun zhen-yu Zhang Qiao, and Li Jing-yan Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems., Tianjin AECC Commercial Aircraft Engine Co.,Ltd. Global Crown Technology Co., Ltd. University of Technology Shanghai, China Beijing China Tianjin, China • Convert the system-level model to a standardized system architecture which • Using EEMD to improve the EMD contains structure and constraint and combine with ARMA to • The system-level model based on forecast the load energy Systems Modeling Language (SysML) consumption. can be exported to Extensible Markup • Taking historical data of a plant's Language (XML) which can rebuild the equipment as a sample. system architecture using Meta model. • Based on EEMD-ARMA prediction In this way, the system architecture Integrated design process of results and error analysis, the becomes a guide for aircraft model- Aircraft and Propulsion prediction accuracy is higher based design and verification. System Algorithm Prediction Results

IEEE ICMA 2020 Conference Digest TA2-5: Cognitive Science and Neural Engineering Methods (II)

Session Chairs: Shuxiang Guo, Kagawa University Qiong Wu, Suzhou University of Science and Technology Online Conference Room 5, UTC+8(Beijing Time): 11:15-12:15, Thursday, 15 October 2020

TA2-5(1) 11:15-11:30 TA1-5(2) 11:30-11:45 The Neural Representation of Room Locations in Visual Orientation Sequence Task enhances A Virtual Building Working Memory in Younger and Older Adults Liwei Sun1, Peter U. Tse2, Chunlin Li1, Xu Zhang1 on ICMA 2020 1School of Biomedical Engineering, Capital Medical University, Beijing, China 2Department of Psychological and Brain Sciences, Dartmouth College, USA Ting Guo, Yanna Ren, Yinghua Yu, Yiyang Yu,Seongyeol Yun, Qiong Wu, Jiajia Yang, Satoshi Takahashi, Yoshimichi Ejima, Jinglong Wu • We found that the hippocampus The Graduate School of Natural Science and Technology, Okayama University, tracks one’s real-time location in a Okayama, Japan building. ➢ The first point our results demonstrated was that • A virtual-reality building was built both younger and older groups had an improving performance between pre and post training. with eight novel objects as ➢ Our second point was that suitable difficulties landmarks. needed to be carefully chosen for different groups. • Participants were actively searching ➢ The third point we would like to address was that for target landmarks in the VR individual factors influenced the results of evaluation tasks and this possibly had to do with building during fMRI scanning. aging effects. Fig. 1 An Example trial of the search task. ➢ One last point we needed to make here was that • Participants’ real-time locations older participants’ group had a higher relative could be decoded from their fMRI growth in accuracy before and after training than data in the hippocampus. that of the younger group’s.

TA2-5(3) 11:45-12:00 TA1-5(4) 12:00-12:15 Development and performance evaluation of automatic A Novel Reduced-Order Fault Detection measuring apparatus for sorting haptic angle stimulus Filtering Design for Singular Continuous-Time Qiong Wu, Yang Liu, Yulong Liu, Jinglong Wu, Ritsu Go Department of Psychology, Suzhou University of Science and Technology, China Markovian Jump Systems Shi Yunling, Peng Xiuyan For early detection of dementia, the aging effect between young College of Automation, Harbin Engineering University people and the elderly was studied by angle chip rearrangement experiments. As a comparison method, an error score was used. Harbin, China However, because it took time and effort to calculate the error score, we studied to develop a device that can automatically output error • This paper proposes a novel full- scores using a computer. The content of the experiment is to touch order and reduced-order fault the angular stimulus with the fingertip and rearrange them in the order of magnitude of the angle. Experiments were conducted for detection filters design method for young people and the elderly. We conducted two types of experiments, a small angular interval and a large angular interval. singular continuous-time Markovian Angles of 20°to 32°were presented at 2°intervals, 20°to jump systems with completely 44°at 4°intervals, angle stimulation at the left end (minimum) and right end (maximum) respectively fixed. As a result of known transition rates. experiments, the elderly people got bigger errors than young people. From this result, aging effect can be confirmed, time can be • Using the idea of congruent shortened compared to the conventional error score calculation method, and it was possible to gather a lot of data easily. Automatic device of tactile transformation to solve the convex Furthermore, in order to make this device one of the methods of angle discrimination optimization problem, the reduced- diagnosing dementia, further improvements will need to be made Residual signals and examined in the future. order fault detection filter is obtained.

IEEE ICMA 2020 Conference Digest TA2-6: Medical, Biomedical and Rehabilitation Systems (II)

Session Chairs: Xueshan Gao, Beijing Institute of Technology Dongdong Bu, Beijing Institute of Technology Online Conference Room 6, UTC+8(Beijing Time): 11:15-12:15, Thursday, 15 October 2020

TA2-6(1) 11:15-11:30 TA2-6(2) 11:30-11:45 Convolution Neural Network(CNN)-based Upper 3D residual dense convolutional network for Limb Action Recognition diagnosis of Alzheimer’s disease and mild Wenyang Gao, Shuxiang Guo, Dongdong Bu cognitive impairment Key Laboratory of Convergence Medical Engineering System and Healthcare Technology, Jiaxi Niu and Xiaoying Tang the Ministry of Industry and Information Technology, Beijing Institute of Technology School of Life Science, Beijing Institute of Technology, Beijing, China • The paper uses lightweight CNN to • We integrated the residual dense effectively classify and predict a connection module into the large number of sEMGs in different maximum information and gradient parts. Since a lightweight network is flow between the layers in the 3D used, low latency can be achieved convolutional network. for control effects. • We developed the 3D-ResdenseNet • Theoretically, a training model with model for classification detection of strong realtime performance and AD, MCI and Normal contrasts. high accuracy can be achieved, and • We used the MRI data of the ADNI it also has a considerable effect in dataset for experiments to prove the Model evaluation index the control process. superior performance of the proposed Confusion matrix of AD/MCI/ model. Normal contrasts (NC)

TA2-6(3) 11:45-12:00 TA2-6(4) 12:00-12:15 Phase synchronization of EEG in the head midline is A Self-care Assistive Robot for the People with associated with sleep-wakefulness stages Lower Extremity Dyskinesia Yanjun Li, Zhi Xu, Yu Zhang, Xianglin Yang and Xiaoying Tang China Astronaut Research and Training Center Juzhong Zhang, Yuyi Chu, Zhisen Wang, Liming Cai and Hongbo Yang Beijing, China Suzhou Institute of Biomedical Engineering and Technology, CAS, Suzhou, China wakefulness deep sleep light sleep REM sleep • Whether the phase-locking value of 0.04 Compared with traditional walking 0.02 EEG on the boundary between left and PDF 0 AIDS, the self-care assistive robot can 0 0.1 0.2 0.3 PLV 0.4 0.5 0.6 0.7 0.8 right brain hemispheres is associated EEG 0.04 assist users to independently complete with sleep stages or not is still 0.02 PDF lifting arm unknown. The purpose of this paper is 0 daily activities, such as getting in and out 0 0.1 0.2 0.3 PLV 0.4 0.5 0.6 0.7 0.8 adjustable seat δ to evaluate the changes of PLV-based 0.04 bed, going to the toilet, rehabilitation 0.02 phase synchronization during different PDF training, outdoor travel and shopping in 0 0 0.1 0.2 0.3 PLV 0.4 0.5 0.6 0.7 0.8 sleep stages. θ electrical control system 0.04 supermarkets without the help of other 0.02 • PLV of EEG on the boundary between PDF people, so that users can basically take 0 0 0.1 0.2 0.3 PLV 0.4 0.5 0.6 0.7 0.8 left and right brain hemispheres α 0.04 increases from awake to sleep. PLV , care of themselves. In particular, the user

θ 0.02 PDF can perform all the operations with one moving chassis step training mechanism PLVα and PLVβ are all highest in deep 0 0 0.1 0.2 0.3 PLV 0.4 0.5 0.6 0.7 0.8 β sleep, representing a high level of The probability density function (PDF) of PLV hand alone, which is very beneficial for phase synchronization in neurons of the for standard EEG bands during wakefulness, patients with hemiplegia. The self-care assistive robot REM sleep, light sleep and deep sleep brain during deep sleep..

IEEE ICMA 2020 Conference Digest TP1-1: Control Theory and Application (III)

Session Chairs: Chao Jia, Tianjin University of Technology Qinxue Pan, Beijing Institute of Technology Online Conference Room 1, UTC+8(Beijing Time): 13:30-15:00, Thursday, 15 October 2020

TP1-1(1) 13:30-13:45 TP1-1(2) 13:45-14:00 RBF-NN based Upper-bound Adaptive learning Time Derivativers of Zhang Dynamics with Han Sliding Mode Control for Grid-Connected Tracking Differentiator Linked and Also Converter Distinguished Duidi Wu, Bo Long, and Yaozong Cheng Xiaoyan Zheng, Yunong Zhang, Liangjie Ming, Min Yang, and Binbin Qiu University of Electronic Science and Technology of China Sun Yat-sen University, Guangzhou, China Chengdu, Sichuan Province, China

Q Q 1 2 • A series of time-derivative solvers, i2 R • Utilized a RBF-SMC U dc C 2 L termed Zhang time derivativers (ZTD) i c Q Q C 1 V method of grid-connected 3 4 g is introduced to estimate the time inverter and proposed an derivatives of a time-varying signal. improved upper bound Upper bound adaptive learning Controller RBF-NN SMC control i1 Output • The general form of the simplest type Middle Input Layer Layer Layer c adaptive learning sliding sign w 1 h1 i ˆ ()t 1 h 2 e  w2 . of ZTD is obtained through . i 1 PWM Driver w . * u n + mode control. h n t − i1 u 0 mathematical induction. • Has the advantage of fast H(s) u 1 Nominal model control • The links and differences between Han response and excellent Block diagram tracking differentiator (HTD) and ZTD tracking performance. Schematic of RBF-SMC are discovered and discussed.

TP1-1(3) 14:00-14:15 TP1-1(4) 14:15-14:30 Research on Feedforward Control Based on Application of Improved PSO-ELM in Auto Robot Dynamics Parameters Identification Insurance Customer Risk Level Prediction Xingmao Shao, Shoujun Wang, Liusong Yang and Nan Liu Xi Chen, Liang Xu*, Xiang Guo，Yalong Wang, Xiang Zhai Tianjin Key Laboratory for Advanced Mechatronic System Design and Intelligent Control, Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems, and Tianjin University of Technology, Tianjin, School of Electrical and Electronic Engineering • The accurate dynamics parameters Operation Sharing Department, PICC Property and Casualty Company Limited are the basis for establishing a dynamics model and the feedforward • Give strong guidance to the auto control scheme effectively improve insurance industry. the error convergence performance. • The analytic hierarchy process • Deduced the basic parameters of the (AHP) weights risk factors. UR5 manipulator. • Particle swarm optimization limit • Established dynamic model of the learning machine fused with golden UR5 manipulator. sine. • Established a feedforward control Auto insurance risk rating mechanism based on dynamics using The 6-DOF UR5 Manipulator index system structure the Simulink platform.

TP1-1(5) 14:30-14:45 TP1-1(6) 14:45-15:00 Discrete sliding mode repetitive control based on parabola approach law Comparative Study of Model Order Reduction for Chao Jia, Xuewei Gao,Jiaqi Wei, and Haocheng He Linear Parameter-Variant Thermal Systems Tianjin Key Laboratory for Control Theory and Applications in Complicated Systems Henrik Zeipel, Tobias Frank, Mark Wielitzka, Tobias Ortmaier Tianjin University of Technology Tianjiin 300384,China Leibniz University Hanover, Institute of Mechatronic Systems

• The discrete sliding mode controller based on parabola approach law can effectively suppress chattering. • Large Scale Thermal Modeling • The discrete sliding mode repetitive • Model Order Reduction of LPV controller based on ideal switching Systems dynamics can solve the problem of • Rational Krylov & Balanced periodic disturbance. Truncation • The convergence characteristics of Permanent magnet Reduction Error Comparision parabola approach law are analyzed. synchronous linear motor

IEEE ICMA 2020 Conference Digest TP1-2: Signal and Image Processing (III)

Session Chairs: Shuoyu Wang, Kochi university Lanyong Zhang, Harbin Engineering University Online Conference Room 2, UTC+8(Beijing Time): 13:30-15:00, Thursday, 15 October 2020

TP1-2(1) 13:30-13:45 TP1-2(2) 13:45-14:00

Semantic Image Segmentation on Snow Towards Industrial Scenario Lane Detection: Driving Scenarios Vision-Based AGV Navigation Methods Shuai Liu, Mingkang Xiong, Weilin Zhong, Huilin Xiong Yayun Lei, Takanori Emaru, Ankit A.Ravankar, Yukinori Kobayashi and Su Wang Shanghai Key Laboratory of Intelligent Sensing and Recognition, Shanghai Jiao Tong University Graduate School of Engineering, Hokkaido University Shanghai, China Sapporo, Japan • Apply lane detection to the AGV • Semantic segmentation in adverse platform in an industrial environment weathers requires specialized datasets. to realize automatic driving of AGV. • A semantic segmentation dataset on • Navigation band detection based on snow driving scenarios is introduced. image processing. • A synthetic snow dataset is created based on conditional generative • Navigation band detection based on adversarial networks. deep learning. • Experiments are conducted to validate • Both of the two methods achieve over The AGV platform and navigation and test segmentation performance. 90% detection accuracy and the latter band in our industrial application improved the detection accuracy by scenario 5.2% than the former.

TP1-2(3) 14:00-14:15 TP1-2(4) 14:15-14:30 Vision-based edge-line Assisted Moving Vehicle CapsNet based on Encoder and Decoder for Detection Object Detection Man Luo, Xin Wang, Hongbin Ma Guanqi Ding, Jing Bai, Hui Lu, Peng Zhang and Xiansheng Qin School of Automation, Beijing University of Technology School of Mechanical Engineering, Northwestern Polytechnical University, Beijing, China Xi’an, China • The CapsNet-V1 models consist • Moving vehicles detection based on of encoder and decoder. the edge-line assisted images cutting. • The classification net is used as encoder to extract multi-class information. • LaneNet based lane detection and Yolo based target detection. • The reconstruction net is regarded as decoder to obtain masks with multi-object position information. • Image cutting and re-splice. Moving vehicles detection • Based on the randomly expanded MNIST dataset, we evaluate the The overall architecture multi-object classification and of CapsNet-V1 reconstruction abilities.

TP1-2(5) 14:30-14:45 TP1-2(6) 14:45-15:00 Based on Siamese Network with Self-Attention A Review of DMRA Based on WT Model for Gait Recognition Hong Wu, Xiaodong Hua School of Electronic Engineering, Heilongjiang University Yu Zhang, Juan Qin*, Lianrong Lv*, Zhiyong Wang Harbin, China School of Electrical and Electronic Engineering Tianjin University of Technology • The principle of wavelet transform Tianjin, China is briefly introduced, and the • Self-Attention module is added to characteristics of other modulation the Siamese network for gait recognition methods are analyzed. recognition. • The research methods of wavelet • In the training stage, companion recognition are summarized. loss is added to supervise the • Something about the application of middle layer. wavelet transform is explained. • Experimental data show that the • Parameter setting problems and the network has improved the average development of wavelet recognition The overview of the proposed Wavelet recognition rate across view on are also given. CASIA-B database. approach for gait recognition

IEEE ICMA 2020 Conference Digest TP1-3: Mobile Robot System (III)

Session Chairs: Aiguo Ming, University of Electro-Communications Ruochen An, Kagawa University Online Conference Room 3, UTC+8(Beijing Time): 13:30-15:00, Thursday, 15 October 2020

TP1-3(1) 13:30-13:45 TP1-3(2) 13:45-14:00 Modeling and Experimental Verification of Research on 3D Reconstruction based on Bionic a New Spherical Underwater Robot Head-eye Cooperation Ruochen An, Shuxiang Guo, Liang Zheng, Tendeng Awa and Wenbo Sui Peng Zhang, Liying Su, Mingyu Cui and Nianqiang Cui Graduate School of Engineering, Kagawa University, Takamatsu, Kagawa, Japan College of Mechanical Engineering and Applied Electronics Technology Beijing University of Technology, Beijing, China

• The model of the extreme positions • Spherical underwater robot. of head-eye coordination is built and • Motion state simulation. the Information Obtained Frequency • Computational fluid (IOF) is obtained. dynamics simulation. • Direct 3D reconstruction with fixed pose pre-processing. • ADAMS simulation. Structure of Proposed Spherical Robot The Result of Node Reconstruction

TP1-3(3) 14:00-14:15 TP1-3(4) 14:15-14:30 Study on the Motion Control System of the Triphibian Robot based on Fuzzy PID Control Path Planning for Mobile Robots Based on Jian Guo1 ,Kaitian Zhang1, Shuxiang Guo1,2, Jigang Xu3 Improved RRT Algorithm 1 Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Intelligent Robot Lei Shao, Hanze Liu, Chao Chen, Dongdong Du, Ji Li and Hongli Liu Laboratory, Tianjin University of Technology, Tianjin, China Tianjin Key Laboratory for Control Theory & Applications in Complicated system 2Intelligent Mechanical Systems Engineering Department Faculty of Engineering, Tianjin, China Kagawa University, Kagawa , Japan 3Qinghai Haidong,810700,China • Use the ant colony algorithm to iterate • A new type of triphibian Generate first-generation robot was designed. and optimize the path obtained by the RRT paths basic RRT algorithm, and finally • Thr o ug h mode l i ng a nd Ant colony algorithm analysis, a controller based on converge to a shortest path. optimizes the first fuzzy PID is designed to • The improved algorithm ensures the generation paths control the robot motion. success rate, greatly reduces the length of Traverse all ants until find the target point • It could be used in line the path inspection, aerial photography, • Reduces the number of nodes in the path Output shortest path military equipment and The prototype of the triphibian robot • Improves the efficiency of path planning. The Structure of cluster attack. Improved Algorithm

TP1-3(5) 14:30-14:45 TP1-3(6) 14:45-15:00 Robot path planning algorithm Optimal Design of Laser SLAM Algorithm Based based on Bi-RRT and Potential Field on RBPF improved resampling technology Xiaoya Tang, Feng Chen Jianyun Ni，Hainan Wang,Yong Qi, Mingyang Zhao Department of Electrical Engineering, Nantong University Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems Nantong Jiangsu ，China Tianjin University of TechnologyTianJin, China • Combine Bi-RRT and improved potential field. • The distribution of proposals is • Introduce the potential field as the optimized,Improve sampling heuristic information of random accuracy trees expansion, and rebuilted the • The minimum variance resampling virtual target points of two-way. method (MSV) is used to ensure • Adopt the goal bias strategy to the diversity of particles after ensure the direction and randomness resampling of random trees expansion. • More accurate map building SLAM Map Building • Verify the effectiveness and The path planning result superiority of the hybrid algorihm.

IEEE ICMA 2020 Conference Digest TP1-4: Modeling, Simulation Techniques and Methodologies (III)

Session Chairs: Liwei Shi, Beijing Institute of Technology Xihuan Hou, Beijing Institute of Technology Online Conference Room 4, UTC+8(Beijing Time): 13:30-15:00, Thursday, 15 October 2020

TP1-4(1) 13:30-13:45 TP1-4(2) 13:45-14:00 Underwater Obstacle Avoiding Trajectory Simulation Analysis of Damage Mechanism of Tracking Approach for Amphibious Suspended Graphene Film Qin Wang, Ying Liu, Xiande Zheng, Guishan Wang, Yong Zhang, Jing Qiu and Guanjun Liu Spherical Robots Science and Technology on Integrated Logistics Support Laboratory National University of Defense Technology Xihuan Hou, Shuxiang Guo, Liwei Shi, Huiming Xing, Zan Li, Debin Xia, Changsha, Hunan Province, China Mugen Zhou and Yu Liu School of Automation, Beijing Institute of Technology Beijing, China • Most graphene films break • This paper solved the problem of when they reach the maximum underwater trajectory tracking for deflection in the triple the Amphibious Spherical Robot. junction generation stage. • Designing a Model Predictive- • The graphene with round based two-layer controller to re- pinhole defect is easier to keep intact during the drying plan trajectory for avoiding The model and force diagram process compared to the obstacles and track a desired when the triple junction is trajectory as accurately as The avoiding obstacle strategy graphene with oval pinhole when the robot is tracking a trajectory formed possible. defect.

TP1-4(3) 14:00-14:15 TP1-4(4) 14:15-14:30 AUV Path Planning Based on Improved RRT and Multi-AUV Circular Formation Sliding Mode Bezier Curve Optimization Control Based on Cyclic Pursuit Juan Li1，2, Chengyu Yang2 Li Juan1，2,Xinnina Chen2 1.Science and Technology on Underwater Vehicle Technology, Harbin Engineering University 1.Science and Technology on Underwater Vehicle Technology, Harbin Engineering University 2.College of Automation Harbin Engineering University 2.College of Automation Harbin Engineering University

• Using the method of image • Using the control algorithm based processing to establish the AUV on cyclic pursuit to enable multiple underwater working environment. AUVs to complete the real-time hunting of moving target in • Improving the traditional RRT circular formation. algorithm. • Designing sliding mode trajectory • local re-planning in unknown tracking controller and proving the environment stability of controller. Trajectories of multiple AUVs • Simulation results show the Path diagram • Verifying the feasibility of the hunting moving target in effectiveness of the improved control algorithm through the circular formation algorithm simulation experiment.

TP1-4(5) 14:30-14:45 TP1-4(6) 14:45-15:00

Mechanical Design and Path Planning of A Research on the structure of low voltage DC distribution Robotic Arm for GIS Pipeline Inspection network based on MMC improvement 1,2 1 2* 1 1 1 Tao Liu , Wensheng Li , Xiangyu Sun , Wei Wang , Yuansheng Fang , Ximeng Zhu , Guodong Zhu1 , Jianfeng Hou2 , Taowei Wang1 ，Xiaobo Huang3 ,Ke Wang1 1 2 2 Liqiang Zhong , Gangfeng Liu and Jie Zhao 1 XJ Group Corporation,Xuji Dadao 1298, Xuchang, China 1 Electric Power Research Institute of Guangdong Power Grid Co. Ltd. 2 State Grid Ji Nan Electrical Power Company,Xizhoudong Road 238, Jinan, China 2 State Key Laboratory of Robotic Technology and System, Harbin Institute of Technology 3 China Southern Power Grid Company Limited Dong Guan Electrical Power Company, Xuji • A robotic arm that can be mounted on a Avenue 38, Xuchang, China • In this paper, the AC-DC-AC low-voltage DC distribution pipeline inspection robot to complete network is studied on the AC-DC side. Firstly, this paper * * detection task in GIS pipeline was degiened. proposed an improved sub-module capacitor voltage balance Q P v control model predictive control (MPC) strategy, which v abc sdq Power sj 功率计算 • The key parameters were analyzed, and the established objective functions for the modular multilevel dq calculation converter (MMC) phase current, circulating current and sub- * i dq structural was optimized. The robotic arm has module capacitor voltage respectively, and this paper selected θ Phase θ dq the best combination of switch states through rolling 锁相环(PLL) locked loop abc a lightweight structure and good mechanical optimization, eliminating the weight coefficient and i* simplifying the traditional MPC strategy.Secondly, by j properties.. 模型预测Analog reducing the number of predicted sub-module switch state predictive c控制器ontroller

combinations, the computational complexity of the algorithm D驱动rive • A research for obstacle avoidance path signal is reduced, and combined with the calculation of other MPC 信号 planning was conducted. Simulation results strategies to verify its correctness.Finally,through the 11-level MMC system built in Matlab/Simulink environment, it is MMC model predictive control flow show that the robotic arm can obtain collision Prototype of the verified that the improved MPC strategy has good steady-state characteristics and dynamic characteristics. free path in GIS pipeline. robotic arm

IEEE ICMA 2020 Conference Digest TP1-5: Cognitive Science and Neural Engineering Methods (III)

Session Chairs: James K. Mills, University of Toronto Chuqiao Lyu, Beijing Institute of Technology Online Conference Room 5, UTC+8(Beijing Time): 13:30-15:00, Thursday, 15 October 2020

TP1-5(1) 13:30-13:45 TP1-5(2) 13:45-14:00 Stereo Matching of Urban Remote Sensing Prediction Model of User Purchase Behavior Image Based on Edge Block Coincidence Rate Based on Machine Learning Hui Ning, Yutong Sun, Zhihui Li, and Xin Zhao Xiang Zhai, Peng Shi, Liang Xu, Yalong Wang, Xi Chen Department of Computer Science and Technology, Harbin Engineering University Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems, and School Harbin, China of Electrical and Electronic Engineering • The first point: Preprocessing of remote Operation Sharing Department,PICC Porperty and Casualty Limited sensing images (extract the SURF Tianjin,china feature points,image registration). • Constructing time series features • The second point: Local motion using feature fusion ideas estimation based on block matching. • The importance of features was • The third point: Displacement values analyzed using random forest clustering based on Mean Shift and post algorithm processing based on MRF. • Based on XGBoost and LightGBM, • The fourth point: Dense matching based a set of user purchase behavior on mutual information( stereo matching prediction models are built method ). Matching result Feature screening

TP1-5(3) 14:00-14:15 TP1-5(4) 14:15-14:30 Research on Dynamic Simulation and Cooperative Control of Augmented Reality and Quick Response Code Smart Ammunition Formation Technology in Engineering Drawing Course Jian Shen, Qingyu Zhu, Lin Xu, Guoguang Chen, Xiaoli Tian, Xiaolong Yan College of Mechatronics Engineering, North University of China, Xiaofei Wang, Chenchen Fei, and Zhaotong Li Tianjin Key Laboratory for Advanced Mechatronic System Design and Intelligent Control, Taiyuan China School of Mechanical Engineering； National Demonstration Center for Experimental Aiming at the dynamic analysis and formation Mechanical and Electrical Engineering Education （Tianjin University of Technology） control of smart ammunition formation, the Tianjin, China formation dynamics model and formation control model are established, and the basic dynamic • AR application and QR code characteristics of smart ammunition formation are application are developed. analyzed by simulation in this paper. By analyzing the three-dimensional trajectory of leader and • Both 3D model and the views follower projectiles, the relative position distance can be seen through camera. between the leader and follower can be steadily close to the expected value. The simulation results • The 3D model can be rotated and show that the dynamic model and algorithm of smart moved by touching the screen. ammunition formation are ideal. In the future, the time and space synchronization of leader and • Models and videos can be follower ammunition should be considered, and the obtained by scanning the QR cooperative dynamic control technology of smart Smart Ammunition Formation code. ammunition formation should be further studied. Image in AR app Video in QR app

TP1-5(5) 14:30-14:45 TP1-5(6) 14:45-15:00 Design of Intelligent Watering Robot Pengfei Lv 1, Shigang Wang 1, Jian Li 1, Xueshan Gao 1,2 Self-adaptive Walking Speed Control on 1Robot Institute of Guangxi University of Science and Technology, Liuzhou, China Underactuated Rimless Wheel 2School of Mechatronical Engineering of Beijing Institute of Technology, Beijing, China Wenxin Lai, Yujia Tian, Shujun Han, Yue Lin,Yongjiang Xue and Juezhu Lai Department of Computer Science and Technology，TIANGONG University • It is proposed to design an Tianjin, China intelligent watering robot to reduce • The walker, URW with torso, is human resources, reduce labor introduced, and its equation of intensity of manual spraying, motion is analyzed. improve the efficiency of watering bonsai, and accomplish the • The Self-adaptive speed control automatic intelligent function of system is proposed by dynamically watering bonsai. planning the gait trajectory and updating control parameters. • Able to automatically follow a • Our control system and its capability predetermined path to water the 3D drawing of robot mechanism A model of an Underactuated bonsai that needs watering. of the disturbance rejection are Rimless Wheel(URW) with torso verified by the simulation.

IEEE ICMA 2020 Conference Digest TP1-6: Medical Robots for Minimal Invasive Surgery (I)

Session Chairs: Hideyuki Sawada, Waseda Univesity Xiaoliang Jin, Kagawa University Online Conference Room 6, UTC+8(Beijing Time): 13:30-15:00, Thursday, 15 October 2020

TP1-6(1) 13:30-13:45 TP1-6(2) 13:45-14:00 A Method for Obtaining Contact Force between Catheter A Two-channel Haptic Force Interface for Tip and Vascular Wall in Master-slave Robotic System Endovascular Robotic Systems 1 1,2,3 3 1 1 Xiaoliang Jin , Shuxiang Guo *, Jian Guo *, Peng Shi , and Dapeng Song 1 1,2 1 1 1Department of Intelligent Mechanical Systems Engineering, Faculty of Engineering and Design. Kagawa Peng Shi , Shuxiang Guo , Xiaoliang Jin and Dapeng Song 1 University, 2217-20 Hayashi-cho, Takamatsu, Kagawa, Japan. Department of Intelligent Mechanical Systems Engineering, Faculty of Engineering and 2Beijing Institute of Technology, Beijing, China. 3Tianjin University of Technology, Tianjin, China. Design, Kagawa University, Takamatsu, Japan 2Key Laboratory of Convergence Medical Engineering System and Healthcare Technology, • The force information of guidewires and The Ministry of Industry and Information Technology, School of life science, Beijing Institute of catheters has a significant affect on the Technology,Beijing, China safety of vascular interventional surgery, especially the contact force between the • In this research, a two-channel haptic catheter tip and vascular wall. The design of force sensor force interface for endovascular robotic • In this paper, a force sensor based on the pressure sensitive rubber is developed to system has been developed. extract the contact force between the tip • With a cylinder-based force generator of catheter and vascular wall. and a MRF-based torque generator, the • The results showed that the design of the presented haptic force interface can force sensor is feasible and effective. the realize linear force feedback and purpose of extracting the contact force is The proposed haptic achieved preliminarily. torque feedback. The contact force detected by the developed force sensor force interface

TP1-6(3) 14:00-14:15 TP1-6(4) 14:15-14:30 Design of a Novel Balloon Catheter Delivery Reciprocating Operation Method for Vascular Mechanism for the Vascular Interventional Intervention Surgery Robot Robotic System Kaidi Wang, Hang Yuan, Yuwen Zeng, Nan Xiao*, Key Laboratory of Convergence Biomedical Engineering System and Healthcare Technology, Hang Yuan, Kaidi Wang, Yuwen Zeng, Nan Xiao* The Ministry of Industry and Information Technology, School of Life Science, Key Laboratory of Convergence Biomedical Engineering System and Healthcare Technology, Beijing Institute of Technology, Haidian District, Beijing 100081, China The Ministry of Industry and Information Technology, School of Life Science, Beijing Institute of Technology, Haidian District, Beijing 100081, China • Proposed the action logic of using the • Proposed a noval balloon catheter delivery joystick to control the vascular Guide wire holder mechanism. interventional surgery robot. Catheter holder • The device uses a conveyor belt to drive • Solved the problem of close-range VISR the balloon catheter, and the conveyor belt coordinated delivery of catheter holder can reduce the damage to the balloon. and guide wire holder. • The pushing force and pushing accuracy of • Realized precise delivery of the device meet the clinical requirements. interventional surgical robots under low-speed conditions vascular intervention surgery robot The schematic view of the device

TP1-6(5) 14:30-14:45 TP1-6(6) 14:45-15:00 Irregular Motion Recognition of Guidewire in A Novel Clamping Mechanism for Vascular Interventional Surgery Circumferential Force Feedback Device of the Youchun Ma1, Shuxiang Guo1.2*, Chuqiao Lyu1, Yue Wang1 Key Laboratory of Convergence Biomedical Engineering System and Healthcare Technology, Vascular Interventional Surgical Robot The Ministry of Industry and Information Technology, School of Life Science, Zhengyang Chen, Shuxiang Guo, Wei Zhou Beijing Institute of Technology, Beijing, China Key Laboratory of Convergence Biomedical Engineering System and Healthcare Technology, • In this paper, transfer learning on The Ministry of Industry and Information Technology, School of Life Science, Beijing Institute of VGG16 is used to train the bouncing Technology, Beijing, China motion data collected by the vascular • A novel clamping mechanism is interventional surgery robot. (a) (b) designed to realize circumferential • By processing, refining and integrating force feedback on the master side of the guidewire bounce data, the interventional surgery robots. accuracy rate of the guidewire bounce • The clamping mechanism can action classification in the test set is accurately reproduce the increased to 93.2%. (c) (d) Training Set Data circumferential force measured from • This paper realizes the accurate (a)(b) is the image before and after the bounce the slave manipulator recognition of the guidewire bounce action; (c) the image is (a)(b) the added image; The Clamping Mechanism (d) is the image after (c) processing action.

IEEE ICMA 2020 Conference Digest TP2-1: Control Theory and Application (IV)

Session Chairs: Songyuan Zhang, Harbin Engineering University Jian Huang, Huazhong University of Science and Technology Online Conference Room 1, UTC+8(Beijing Time): 15:30-17:00, Thursday, 15 October 2020

TP2-1(1) 15:30-15:45 TP2-1(2) 15:45-16:00

Rope-hook recovery system of fixed wing UAV Research of Hydrodynamic Coefficients based on FFT prediction algorithm and adaptive Identification for Submarine in Vertical Motions fuzzy PID control Based on APSO MingXi Zhang, Qi Li , LingLing Chu, XinTian Du, Feng Gu, YuQing He Ping Zhang, Changbo Liu, Kun Hu, Jianhua Zhang State Key Laboratory of Robotics., Shenyang Institute of Automation, Department of Operation and Command, PLA Naval Submarine Academy

Chinese Academy of Sciences, China Encoder Qingdao, China Each joint - Tk-1 Mathematical Desired T model of position Fuzzy-PID k Motor • A control strategy combining recycling controller - FFT Signal mechanism prediction Execute response Dynamic Relative pose data differential adaptive prediction algorithm GPS + • Hydrodynamic coefficients with fuzzy PID to construct the Wave interference Aerodynamic interference • An adaptive particle swarm rope-hook recovery system for optimization algorithm the recovery of fixed wing • Equations in vertical plane UAVs was proposed. • Coefficient identification • Kalman filter is used to build an adaptive disturbance Curve of coefficients prediction algorithm; The simulation test platform identification about Zw

TP2-1(3) 16:00-16:15 TP2-1(4) 16:15-16:30

Modeling and Simulation of Seeker Servo System Position Sensorless Control of Switched Reluctance Motor Based on Full-bridge Power Haiyan Qiao , Hao Meng,Fanqiang Meng, Zhongbin Li and Zhenhua Zhang Harbin Engineering University Converter Harbin,China Chenguang Qiu ，Yueming Guan，Yan Liu and Xudong Fu Key Laboratory of dual medium power technology Hebei East China Institute of Optoelectronic Integrated Devices Handan, China Bengbu, Anhui, China • This paper proposes and investigates • The composition and function of the a high frequency pulse injection Orientation Orientation seeker servo system. frame motor method combined with five phase Gyroscope full bridge power converte. • The mathematical models of DC Load torque motor, gyro, resolver and Main support • The proposed method has the power amplifier. Pitch angle sensor Pitch motor property of improveing the system

Orientation • speed negative feedback double Pitch frame stability, controling the design cost . angle sensor closed-loop correction system. • Simulation and experimental results Structure of a seeker servo have verified the efficacy of the system proposed method. Motor hardware control platform

TP2-1(5) 16:30-16:45 TP2-1(6) 16:45-17:00 Neural network Approximation-based Nonlinear Research on Real-time Simulation Method of Model Predictive Control for Station-keeping of A Vascular Interventional Surgery Based on Multi-vectored Propeller Airship Model Order Reduction Yubin Wen, Qunhua Pan, Li Chen, and Dengping Duan Baofeng Gao1, Lamei Shang1 China Aeronautical Radio Electronics Research Institute, 1 Key Laboratory of Convergence Biomedical Engineering System and Healthcare Technology, Aviation Industry Corporation of China The Ministry of Industry and Information Technology, School of Life Science, Beijing Institute of Technology, Shanghai, China • A Neural network approximation- • Use model reduction method based nonlinear model predictive to speed up blood vessel control is proposed, and the stability simulation rate. analysis is given. • Use multi-threading and • The station-keeping of the multi- CUDA to speed up the vectored propeller airship is simulation speed of the investigated. system. • Simulation results have verified the The Multi-vectored Propeller Airship The simulated frame rate of the system in different situations efficacy of the proposed method.

IEEE ICMA 2020 Conference Digest TP2-2: Signal and Image Processing (IV)

Session Chairs: Enzeng Dong, Tianjin University of Technology Chuqiao Lyu, Beijing Institute of Technology Online Conference Room 2, UTC+8(Beijing Time): 15:30-17:00, Thursday, 15 October 2020

TP2-2(1) 15:30-15:45 TP2-2(2) 15:45-16:00 Gait recognition in real environment using gait Design of A Novel Virtual Sensor for the energy image generated by Mask R-CNN Vascular Interventional Robotic System Shota Inui, Fuji Ren, Shun Nishide and Xin Kang Jian Guo1, Han Zhao 1 Shuxiang Guo1,2 Faculty of Engineering, Tokushima University, Japan 1Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Intelligent Robot Laboratory • The purpose of research is to improve Tianjin University of Technology, Tianjin, China the accuracy of gait recognition. 2.Department of Intelligent Mechanical Systems Engineering Faculty of Engineering • Use Mask R-CNN to create gait Kagawa University, Japan silhouettes with noise removal such as • The virtual sensor was proposed to obtain more accurate force information and shadows and lighting variations. solve the problems of traditional sensors in vascular interventional surgery. • Also possible to remove the area of • The system identification technology was used to establish the kinematic model of the possessed object. the catheter. • Use CNN enhanced by Batch • The experiment is to study the Normalization to identify people. measurement of collision force by • Confirmed effectiveness by simulating real blood vessels with glass vessels, and the results show conducting two types of walking good measurability. experiments with and without bag. Mask R-CNN The experimental platform diagram

TP2-2(3) 16:00-16:15 TP2-2(4) 16:15-16:30 Application of Improved Canny Algorithm in An Automatic Object Detection and Tracking Image of Surgical Instruments Method Based on Video Surveillance Hongli Liu ,Kuan Lu , Lei Shao and Ji Li Enzeng Dong, Yue Zhang, and Shengzhi Du Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems., Tianjin School of Electrical and Electronic Engineering,Tianjin University of Technology University of Technology TianJin, China Tianjin, China

• Proposing an improved Canny • Basic theory of object detection algorithm with adaptive high and GMM bbaacckkggrroouunndd HOG ffeeaattuurree eexxttrraaccttiioonn T T

r modeling low thresholds.

r modeling a a

and tracking algorithms f f f f i i c c

v v i i d d Deetteeccttiioonn e e o o

• Improved algorithm used in edge

i model n model • Design automatic detection and n p p u u

t output t output Extract moving objects Extract moving objects SVM classifier detection of surgical instrument tracking mechanism area(ROI) SVM classifier area(ROI) images. • Design decision module ROI area extraction Specific category recognition • Result analysis • Experimental results show that the Detection model flowchart improved algorithm can better retain the edge information of surgical instrument images. Gray Scale Image and Edge Detection Image

TP2-2(5) 16:30-16:45 TP2-2(6) 16:45-17:00 Detection of road surface crack based on PYNQ UAV Remote Sensing Image Mosaic Zhang Yuhan, Qin Juan*,Guo Zhiling, Jiang Kuncheng, Cai Shiyuan Technology Combined with Improved School of Electrical and Electronic Engineering Tianjin University of Technology SPHP Algorithm No.391 BinShuiXiDao Road, Tianjin, China Wang Sining,LengJianwei Tianjin Key Laboratory of Complex Control Theory &ApplicationTianjin University of • Aim at the problem of road surface Technology image crack recognition by using Tianjin, China embedded camera. • In this paper, an algorithm with good • Using a lower-cost embedded robustness to topography is presented. camera to capture photos • Compared with traditional SIFT algorithm • The crack profile is marked by the and unimproved SPHP algorithm, the FAST feature point and use Python algorithm presented in this paper has on Zynq(PYNQ) to process. obvious improvement in time consuming • It makes the identification system and image stitching effect. more integrated and portable. Image splicing process Recognition result by using PYNQ

IEEE ICMA 2020 Conference Digest TP2-3: Mobile Robot System (IV)

Session Chairs: Liang Zheng, Kagawa University Xiufen Ye, Harbin Engineering University Online Conference Room 3, UTC+8(Beijing Time): 15:30-17:00, Thursday, 15 October 2020

TP2-3(1) 15:30-15:45 TP2-3(2) 15:45-16:00 Quadrotor Location-based Target Hunting Strategy A Robotic Wheel Locally Transforming the for Multiple Amphibious Spherical Robots Diameter for the Locomotion on Rough Terrain Debin Xia, Shuxiang Guo, Liwei Shi, Huiming Xing, Xihuan Hou, Zan Li, Mugen Zhou Naoki Moriya, Hiroki Shigemune and Hideyuki Sawada Key Laboratory of Convergence Medical Engineering System and Healthcare Technology, Department of Applied Physics, Waseda University, Tokyo, Japan the Ministry of Industry and Information Technology, Beijing Institute of Technology

• We develop a robotic wheel, which • In order to realize the underwater functions locally transforms its diameter for the of amphibious spherical robots and mobility on uneven terrains. complete the task in water, a multi-robot • Extension and returning mechanism cooperative hunting strategy for target is is realized by rack & pinion structure. proposed. • Trajectory of the spoke tip is verified • In the target hunting process, multiple by the simulation using MATLAB. robots converge uniformly when they are hunting target, and they form a triangle • Locomotion on flat ground, climbing formation. over a step, and braking actions are controlled by the spoke articulation. Concept of locomotion, and • The simulation results show the the assembled robotic wheel effectiveness of this target hunting method.The target hunting diagram

TP2-3(3) 16:00-16:15 TP2-3(4) 16:15-16:30 Study on Map Construction of Spherical Robot Multi-functional Smart Car Based on Wireless based on Statistical Filtering Communication Technology Jian Guo1 ,Xiangyu Chen1, Shuxiang Guo1,2, Jigang Xu3 1Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Biomedical Robot Laboratory, Junchao Zhu, Huifeng Cao,Baofeng Zhang, Yunlong Xing and Chang Jia Tianjin University of Technology, Tianjin, China Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems., Tianjin Intelligent Mechanical Systems Engineering Department Faculty of Engineering, Kagawa University, Kagawa , Japan University of Technology 2Intelligent Mechanical Systems Engineering Department Faculty of Engineering, Tianjin, China Kagawa University, Kagawa, Japan 3 Qinghai Haidong,810700,China • A new method is proposed to collect • ORB_SLAM2 is applied to mobile information on polluted environment robots by using intelligent vehicle carrying • The volume of point cloud image is information collection system. reduced by means of filtering • Collect a variety of harmful gas • The point cloud map is converted concentrations into octree map for robot navigation • Experiments verify that each sensor has high accuracy and is suitable for Information Acquisition System The Octree map detecting harmful gases in a polluted environment.

TP2-3(5) 16:30-16:45 TP2-3(6) 16:45-17:00 Design of Cascade Control Framework for Dynamic Sampling RRT for Improved ROV Control and Simulation Performance in Large Environments Elmkaiel Ghadeer, Serebrenny Vladimir Valeryvich 1,2 1,2 3 1,4 Changlin Qiu , Shihan Kong , Chao zhou , and Junzhi Yu Mechatronics and Robotics Dep., Bauman Moscow State Technical University 1 State Key Lab Management and Control for Complex Systems, Institute of Automation, CAS Moscow, Russia 2 School of Artificial Intelligence, University of Chinese Academy of Sciences 3 Naval Research Academy 4 College of Engineering, Peking University • RRT is one of the main Random Beijing, China Sampling path planning algorithm. • A cascade control framework is Interference • Dynamic Sampling RRT is an proposed in class of ROVs to Cascade control TP + improved version with better F + S enhance the capacity of P Fuzzy ROV performance in large and complicated disturbance resistance. Controller TV environments. • The cascade control framework is Measurement V PID Neural V Velocity • The main principle is to dynamically Network Feedback S composed of fuzzy controller and Controller limit the sampling space around the P F Position PID neural network controller Feedback generated tree. Dynamic Sampling RRT • The contrast simulations are • New random points inside the designed and the results are The structure of cascade control framework dynamic space will have much higher satisfactory chance to go through narrow passages.

IEEE ICMA 2020 Conference Digest TP2-4: Modeling, Simulation Techniques and Methodologies (IV)

Session Chairs: Hideyuki Hirata, Kagawa University Zixu Wang, Kagawa University Online Conference Room 4, UTC+8(Beijing Time): 15:30-17:00, Thursday, 15 October 2020

TP2-4(1) 15:30-15:45 TP2-4(2) 15:45-16:00 Modeling and Simulation of the Drug Delivery Task Allocation of Rescue Robots based on Path Function for a Magnetic Driven Capsule Robot Optimization in Chemical Disaster Environment Zixu Wang, Shuxiang Guo and Wei Wei Ziyan Lin, Yingqiu Xu and Yingzi Tan Graduate School of Engineering, Kagawa University, Takamatsu, Kagawa, Japan School of Mechanical Engineering, Southeast University Nanjing, China • Chemical disaster events are characterized by suddenness, danger and variability and the • Magnetic driven capsule Rescue robots are inefficient and unsafe . robot. • Combining the characteristics of the • Magnetic electronic field. environment, a task framework path is • Drug delivery. proposed. The improved path planning • Ansys Fluent simulation. method and task allocation method are proposed • 6-DOF interial measurement Chemical disaster unit. • The simulation results show that the methods can be well adapted to the chemical disaster Drug delivery simulation environment

TP2-4(3) 16:00-16:15 TP2-4(4) 16:15-16:30

Evaluation Performance of the Magnetic Research on AUV Path Planning Based on Rotational Field for Magnetically Actuated Improved Ant Colony Algorithm Microrobot System Li Juan1，2, Huixin Wang2 Qiang Fu, Zhuocong Cai,Jian Guo, and Shuxiang Guo 1.Science and Technology on Underwater Vehicle Technology, Harbin Engineering University Tianjin Key Laboratory for Control Theory & Application in Complicated Systems and 2.College of Automation Harbin Engineering University Biomedical Robot Laboratory., Tianjin University of Technology Tianjin, China • Using the grid method to establish • Construction of electromagnetic the AUV working environment. simulation model to evaluate the platform of magnetic drive capsule • Improving the traditional ant robot. colony algorithm. • Use Electromagnetic simulation to • Optimize the path second time. measure torque of the magnetic • Simulation results show the capsule robot. effectiveness of the improved • Achieve more accurate control of algorithm Path comparison diagram robot motion state. The Magnetic flux density of coils

TP2-4(5) 16:30-16:45 TP2-4(6) 16:45-17:00 Mitigation of Voltage Oscillation during Turn-off Period of the Silicon Carbide (SiC) MOSFET Strength Simulation and Test Analysis on Metro Zhaoyi Liu1, and Hui Li2 Bogie Frame of High speed And Heavy Load 1 China Electronics Technology Group Corporation, Beijing, China Shuo Zhang,Ruizhen Dai,Jingwei Chen,Chensheng Li, Xiuyu Sun, and Liangjie Li 2 University of Electronic Science and Technology of China Bogie Technology Center, CRRC Tangshan Co., Ltd • Mechanism of the switching Tangshan, Hebei Province, China oscillation during turn-off transition • Introduction of the SiC MOSFET. • Design structure of frame • Impact of decoupling capacitance and snubber resistance on turn-off • Strength simulation analysis switching oscillation. • Strength test analysis • Determination of values of • Conclusions decoupling capacitance and snubber

resistance. Switching waveforms with and without • Simulation and Experiment. decoupling capacitance and snubber resistance Research Process of Bogie Frame

IEEE ICMA 2020 Conference Digest TP2-5: Cognitive Science and Neural Engineering Methods (IV)

Session Chairs: Xianqiang Bao, Beijing Institute of Technology Linshuai Zhang, Chengdu University of Information and Technology Online Conference Room 5, UTC+8(Beijing Time): 15:30-17:00, Thursday, 15 October 2020

TP2-5(1) 15:30-15:45 TP2-5(2) 15:45-16:00 Shape Fitting and Morphometric Analysis of Zhang Equivalency Originating from Neural Open Curves Based on Discrete Cosine Dynamics for Time-Varying Problems Transform Solving Shengmin Zhou , Bingjue Li Yunong Zhang, Liangjie Ming, Min Yang, Binbin Qiu, and Zhonghua Li School of Mechanical Engineering, Southeast University, Nanjing, China Sun Yat-sen University, Guangzhou, China Corresponding Author: [email protected] • Decrete cosine transform (DCT) is employed to fitting • This paper considers ZE of inequality 2D/3D open curves; type. • Nine groups of 90 leaves are • The order-1, order-2, and order-3 ZE fitted with DCT, analyzed by general formulas of inequality type stepwise FDA. are proposed and further discussed. • Cochlear, lambdoidal and • This paper proposes the ZE formulas nasal-aperture 3D curves in of time-varying inequality problems human skull are fitted . in four different situations, Fitting plot and canonical plot Future application

TP2-5(3) 16:00-16:15 TP2-5(4) 16:15-16:30 Continuous ZNN Models for Computation of Numerical Computation, Symbolic Computation Time-Varying Eigenvalues and Corresponding and Result Analysis of Jordan Decomposition of Eigenvectors Time-Varying Matrices Min Yang, Chuming Li, Frank Uhlig, Haifeng Hu, Yunong Zhang, and Binbin Qiu Zhuoheng Zhen, Yunong Zhang, Xiao Liu, Yihong Ling, and Min Yang Sun Yat-sen University, Guangzhou, China Sun Yat-sen University, Guangzhou, China • A continuous ZNN model is proposed • The Jordan decomposition of constant to compute the eigenvalues and and time-varying matrices is studied. eigenvectors. • The numerical computation and • The convergence and robustness of the symbolic computation of Jordan proposed ZNN model are analyzed. decomposition are studied . • numerical experiments substantiate the • The relationship between the error and efficacy of the ZNN model. sampling interval of the Jordan decomposition of constant and time- • An extensional ZNN model is varying matrices is obtained by developed to solve all eigenvalues and numerical experiments. eigenvectors in real time. EigenvaluesTracking Trajectories Errors of numerical and symbolic computation

TP2-5(5) 16:30-16:45 TP2-5(6) 16:45-17:00 Research on Image Edge Detection Method Constant current and constant voltage output Based on Improved Robinson Operator and Otsu WPT system based on LCC compensation Algorithm network Yankai Shi, Kun Li* , Haibo Zhao,Changsong Wang, Xin Li, Yu Wang, Chenglong Yu,Taifeng Hao Ren, Jie Sun and Wei Yuan Zhang, Bin Hu and Ziyu Wu Electronic Engieering Research Lab., Tianjin University of Technology Tianjin University of Technology;Tianjin 712 Communication & Broadcasting Co;Tianjin Tianjin, China institute of power source. • Robinson edge detection operator and their • A LCC-S circuit is used for constant 7 6.5956.466 6.743 improved algorithm, all of which will 6.711 6.467 6.7 6 6.345 voltage output and a LCC-LCC circuit 5.915 involve the determination of the optimal 5.393 5

is applied for constant current output. 4.402 threshold. 4 4.036

• The compensation strategy can 3 2.724

• By improving the classical maximum inter- current(A) Load stabilize the system output by 2 class variance method, a new threshold 1.34 1 After circuit compensation 1.076 0.817 changing the component parameters in Before circuit compensation 0.538 segmentation algorithm is proposed. 0 0.277 0.6 0.5 0.4 0.3 0.2 0.1 0.0 the LCC network, when the Coupling coefficient k • The proposed new algorithm improves the transmission distance changes. speed and accuracy of edge detection Improved Algorithm Simulation of the compared with classical Otsu methods. compensation

IEEE ICMA 2020 Conference Digest TP2-6: Medical Robots for Minimal Invasive Surgery (II)

Session Chairs: Wei Zhou, Beijing Institute of Technology Qiang Huang, Beijing Institute of Technology Online Conference Room 6, UTC+8(Beijing Time): 15:30-17:00, Thursday, 15 October 2020

TP2-6(1) 15:30-15:45 TP2-6(2) 15:45-16:00

A Novel Axial Force Feedback Device for the A CNN-based Method for Guidewire Tip Collisions Master Manipulator of the Vascular Interventional Detection in Vascular Interventional Surgery Chuqiao Lyu, Shuxiang Guo*, Youchun Ma, Yue Wang Surgical Robot Key Laboratory of Convergence Biomedical Engineering System and Healthcare Technology, Wei Zhou, Shuxiang Guo, Zhengyang Chen Beijing Institute of Technology, Beijing, China Key Laboratory of Convergence Biomedical Engineering System and Healthcare Technology, • Many iatrogenic vascular The Ministry of Industry and Information Technology, Beijing Institute of Technology. injuries come from • A novel axil force feedback device with surgeons’ risky actions non-contact displacement measure in vascular interventional equipment is designed. surgery. • The design based on electromagnetics • In this paper, three CNN and the working principle are models are trained to elaborated in detail. identify the collisions between the guidewire • The calibration experiments of non- The whole structure of the tip and the vascular wall. contact displacement measure axial force feedback device The CNN Model for Guidewire Tip Collisions Detection equipment is done • Our model got more than 95.9% accuracy and made it possible to observe the guidewire tip frame by frame during the operation.

TP2-6(3) 16:00-16:15 TP2-6(4) 16:15-16:30 Recognition and Filtering of Tremor Signals A Novel Remote Center of Motion Algorithm for Vascular Interventional Surgical Robot Based on 3D Force Dragging for Minimally Liuqing Zhang, Shuxiang Guo, Cheng Yang, Rui Shen Invasive Surgical Robot Beijing Institute of Technology Jinqiang Zhao1,2,3,Yuanyuan Zhou2,3,4,Mingquan Guo2,3,4,Tao Yu2,3,4, Hua Zhang1 and Hao • Accurately identify the tremor Liu2,3,4* signal in the signal through 1.Mechatronic Engineering, Shenyang Ligong University, Shenyang, China. 2.State Key support vector machine . Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang, China. 3.Institutes for Robotics and Intelligent Manufacturing, Chinese Academy of • Use average filter to process the Sciences, Shenyang, China. 4.Key Laboratory of Minimally Invasive Surgical Robot, Liaoning tremor signals. Province, China • The system can more accurately • Porpose a novel single-hole minimally identify the jitter signal in the invasive surgical robot suspension arm z original signal, and can perform configuration which has 6 DOFs. effective smoothing while O y • Put forward a corresponding RCM x retaining other normal and smooth algorithm based on 3D force dragging. motion signals. The System • Use ROS and Matlab to do simulation . Remote Center of Motion

TP2-6(5) 16:30-16:45 TP2-6(6) 16:45-17:00 Master-Slave mapping method for three-axis Target Tracking Algorithm for Template intersection laparoscope arm Transformation Jing Yang, Zejie Han, Yang Zhang, Lingyan Jin, Ming Hu based on Template Matching Faculty of mechanical engineering & Automation, Zhejiang Sci-Tech University Hangzhou, China Lanyong Zhang, Pingwei Zhang School of Automation，Harbin Engineering University

• Designed the mechanism of the Joint 6 Joint 7 Harbin, China Joint 2 Joint 3 Joint 4 manipulator with a new type of three- • A template transformation Joint 1 axis intersection remote center matching algorithm. Joint 8 structure. mobile platform Joint 5 • Update the template size and • Established a master-slave consistent Laparoscope template content to improve the motion mapping algorithm based on accuracy of target tracking. window coordinate system. • The tracking recognition rate is Original Template Matching • Simulated and verified the master- The three-axis intersection increased to more than 99.5%, Algorithm VS Template slave motion algorithm by Geomagic laparoscope arm and the algorithm delay is Transformation Marching Touch master hand. controlled within 70ms. Algorithm

Friday October 16, 2020

Morning Sessions

FA1-1 Neuro, Fuzzy, and Intelligent Control (I) FA1-2 Surgical Robot System FA1-3 Digital Design and Manufacture FA1-4 Humanoid Robots FA1-5 Rotor Dynamics, Vibration Analysis and Vibration Control FA2-1 Neuro, Fuzzy, and Intelligent Control (II) FA2-2 Soft Robotics FA2-3 Sensor Networks, Distributed Sensor Systems FA2-4 Human-System Interaction and Interface FA2-5 Control Theory and Application (V)

IEEE ICMA 2020 Conference Digest FA1-1: Neuro, Fuzzy, and Intelligent Control (I)

Session Chairs: Hongjun Wang, Tianjin University of Technology Jin Guo, Beijing Institute of Technology Online Conference Room 1, UTC+8(Beijing Time): 8:30-10:00, Friday, 16 October 2020

FA1-1(1) 8:30-8:45 FA1-1(2) 8:45-9:00 EMG signal Analysis and Identification of Human Attention-based Bi-LSTM-CRF Network for Calf Muscles based on Walking on Different Emotion Cause Extraction in Texts Liyuan Liang1,2, Xiaodong Ji1,and Fuji Ren2 Slope Road 1.School of Information Science and Technology, Nantong University, Junyao Wang, Tong Kang, Zhaoyang Li and Yuehong Dai Nantong, China University of Electronic Science and Technology of China (UESTC) 2.Graduate School of Advanced Technology and Science,University of Tokushima,Tokushima, Chengdu, China Japan • Using EMG signal acquisition platform •We design emotion-clause attention and position-clause attention to measure the EMG signal of calf for a better representation. muscle. •We consider the ECE task as a document-level clause labeling • The feature information of EMG signals problem. was obtained by time and frequency •The utilization of CRF can domain methods. capture the causality between • Seven kinds of road conditions are clauses in a document. identified by BP network. The framework of our model EMG Signal Acquisition Platform

FA1-1(3) 9:00-9:15 FA1-1(4) 9:15-9:30 Optimal Design of Aviation High Power Short-term Photovoltaic Power Prediction Permanent Magnet Motor Based on Particle based on DE-GWO-LSTM Swarm Optimization-Genetic Algorithm Hui Zhao, Zhili Zhao, Hongjun Wang and Youjun Yue Sen Wang;Yanyang Dang;Bai Yang;Zeyu Huang;Yujia Li;Jieqiu Bao;Hongkui Yan; Tianjin Key Laboratory for Control Theory and Applications in Complicated Yan Bao;Di Tang System, Tianjin University of Technology, Tianjin , China Shenyang Institute Of Engineering,TBEA CO.,LTD. • The background and significance Shenyang,China of photovoltaic power generation • Basically elaborate the establishment of the theory and analysis process of electromagnetic • Principle of LSTM network field of high-power permanent magnet motor. • Improved Gray wolf optimization • This article complete the preliminary design of algorithm permanent magnet motor and electromagnetic • Photovoltaic output prediction field simulation analysis. model based on DE-GWO-LSTM • Based on particle swarm genetic algorithm, the • Simulation results and model Model prediction results multiphysics coupling optimization design of analysis high-power permanent magnet motor is Permanent Magnet Motor completed.

FA1-1(5) 9:30-9:45 FA1-1(6) 9:45-10:00 Research on Abnormal Power Fault Type identification in Transmission Lines Consumption Detection Technology Based based on VMD-FE and IGWO-KELM on Decision Tree and Improved SVM Youjun Yue, Lixin Li, Hongjun Wang, Hui Zhao Tianjin Key Laboratory of Complex System Control Theory and Application, Tianjin University Hongjun Wang，Zhiqi Li， Hui Zhao，Youjun Yue， of Technology, Tianjin, China Tianjin Key Laboratory for Control Theory and Applications in Complicated System, Tianjin University of Technology, Tianjin , China • Use VMD to decompose the fault • Combine (CNN) and GAN) for signal to modal components data generation sequence current • Calculate the fuzzy entropy value • DTW-based similarity of each modal component to form measurement a sample library • Improved support vector machine • Combined with differential (SVM) model based on decision evolution algorithm to improve tree (DT) detects user data. the gray wolf algorithm and • When classifying, combine optimize KELM Changes in the proportion of • Use an improved KELM to improved support vector machine Identification results and KNN for classification training samples and accuracy classify fault features

IEEE ICMA 2020 Conference Digest FA1-2: Surgical Robot System

Session Chairs: Chaoyang Shi, Tianjin University Xiaoliang Jin, Kagawa University Online Conference Room 2, UTC+8(Beijing Time): 8:30-10:00, Friday, 16 October 2020

FA1-2(1) 8:30-8:45 FA1-2(2) 8:45-9:00

Design and Experimental Validation of a Master- A Novel Ultrasonic Scalpel Rod with Multi-Stage slave Robotic System for Flexible Ureteroscopy Gain and Minification Structures for Minimally Jianchang Zhao1,2, Shuxin Wang1,2, Liang Cui3, Jianmin Li1,2 * and Chaoyang Shi1,2* Invasive Surgery 1.Key Laboratory of Mechanism Theory and Equipment Design of Ministry of Education Jinhua Li1, Xinyu Dong1, Shuxin Wang1, Zhicheng Guo1, Guokai Zhang2, Chaoyang Shi1* 2. School of Mechanical Engineering, Tianjin University, Nankai District, Tianjin, 300072, China Key Laboratory of Mechanism Theory and Equipment Design of Ministry of Education, Tianjin 3. Civil Aviation General Hospital, Beijing, 100123, China University, Tianjin, China • This paper presents a novel robotic system for flexible ureteroscopy. • The system is configured as a master-slave mode with commercial • This paper presents a novel ultrasonic scalpel rod. endoscope to perform ureteroscopy with more accurate and • The vibration amplitudes are first suppressed to reduce the friction comfortable operation. energy loss and then are enlarged at the distal end for cutting. • The tracking response experiments and animal experiment have been • The FEM-based simulations have been performed. performed to verify the performance of the system. • The liver cutting experiment have been performed.

Master-slave robotic system for flexible ureteroscopy Schematic diagram of the proposed ultrasonic scalpel rod

FA1-2(3) 9:00-9:15 FA1-2(4) 9:15-9:30 Marker-Based Shape Estimation of a Continuum Operation Comfort Analysis for the Master Manipulator Using Binocular Vision and Its Error Manipulator of the Flexible Ureteroscopy Compensation Robot Based on Joint Torque Jinhua Li1, Yanan Sun1, He Su1, Guokai Zhang2, Chaoyang Shi1* 1.Key Lab for Mechanism Theory and Equipment Design of Ministry of Education Chenglong Wang1, Jianchang Zhao1, Xiangyu Luo1, Guokai Zhang2, Jianmin Li1* and Chaoyang Shi1* School of Mechanical Engineering, Tianjin University, Nankai District, Tianjin, 300072, China 1. Key Laboratory of Mechanism Theory and Equipment Design of Ministry of Education 2.Department of Informatics, King’s College London, UK School of Mechanical Engineering, Tianjin University, Tianjin, China Corresponding author: [email protected] 2. Department of Informatics, King’s College London, UK • This paper proposes a maker-based method • Proposing an evaluation model of for shape estimation of a flexible upper limb operating comfort based manipulator using binocular vision. on joint torque for FURS robot. • This paper proposes an error compensation • Establishing the simplified model method based on the OPELM learning of human upper limb. algorithm is also proposed to improve its • Both the muscle fatigue experiments The workflow of the proposed shape accuracy. estimation method using the binocular vision and RULA simulation results verify • Experiments verified that the accuracy of the technique. the reliability of the comfort reconstruction was improved by about 70%. evaluation model.

FA1-2(5) 9:30-9:45 FA1-2(6) 9:45-10:00 Ergonomic evaluation between laparoscopic and Cross-Membrane Penetration to Nucleus of robotic surgery based on EMG Adherent Cells Using Micropipettes Made by Shunkang Niu, Kui Jin, Zhenxuan Hu, Chong Zhang, Di Zhao, Yuan Xing Borosilicate Glass and Quartz Tianjin University, Key Laboratory of Mechanism Theory and Equipment Design of Ministry of Education, Tianjin 300354, China; Xinyue Xu, Xiaoming Liu, Yuqing Lin, Pengyun Li, Fengyu Liu, Xiaoqing Tang, Qiang Huang and Tatsuo Arai • The study compared the ergonomic Beijing Advanced Innovation Center for Intelligent Robots and Systems, Beijing Institute of characteristics of laparoscopic and Technology robotic surgery Beijing, China • The iEMG and first-order fitting • Direct access to the intracellular environment slope of MPF were used for provides an ideal way to gain insight into the comparative analysis. dynamic processes of living cells. • Flexor carpi radialis and • Real nucleus puncture experiments were brachioradialis tend to be more used, conducted on HELA and C2C12 cells. and medial deltoid was less used in • Puncture angle and puncture height are the robotic surgery .There was no adjusted for different cells. significant difference in muscle The “MicroHand S” robot systems • A quartz pipette is an alternative tool suitable Nucleus Penetration of fatigue between two method. for cell or nucleus penetration. Adherent cells

IEEE ICMA 2020 Conference Digest FA1-3: Digital Design and Manufacture

Session Chairs: Qiang Fu, Tianjin University of Technology He Yin, Beijing Institute of Technology Online Conference Room 3, UTC+8(Beijing Time): 8:30-10:00, Friday, 16 October 2020

FA1-3(1) 8:30-8:45 FA1-3(2) 8:45-9:00 An Evaluation Framework to Assess Start-up Stage for an Electric Field Assisted Autonomous Navigation Linked to Fused Deposition Environment Complexity Xu Ruihan1, Bao Weijie1, Yin Tuo1, Wang Zhihai1, Wang Yaohong2 1 Key Laboratory of Optoelectronics Technology, Beijing University of Technology, Beijing Ermacora, Sartori, Rovasenda, Pei, Yu 100124, China Institute of Sensing and Navigation 2 Center for Applied Mathematics, Tianjin University, Tianjin 300072, China Shanghai Jiao Tong University 800 Dongchuan road, 200240, Shanghai, China • Vi is the initial liquid volume at which the nozzle voltage is applied. For small Vi, the • In this work we present a framework for the assessment of autonomous navigation performance referenced to the meniscus deforms into cone shape, and a jet environmental complexity in which the robot operates. much thinner than the nozzle diameter • The framework consists in selecting a set of metrics, proposing guidelines and comparing in the same test appears. scenario theoretical optimal and measured trajectory performances while linking them to the environment • For large Vi, the meniscus exhibits a complexity. spindle shape when it touches the substrate. • Supporting the framework, two innovative algorithms are employed: one for trajectory planning and the other to • It is suggested that for high resolution assess the environment complexity. • Initial results show the applicability of this approach. printing, ramping the voltage at small Vi is preferable. Phenomenon at different start-up volumes

FA1-3(3) 9:00-9:15 FA1-3(4) 9:15-9:30 Path Planning of Tube-Sheet Girth Welding by Design and Research of Intelligent Air Purifier Discrete Firefly Algorithm System Jianhua Ma, Jianyi Kong, Xingdong Wang, Guangming Zou, Yu Huang and Xiufeng Liu Yining Wang, Wenbo Wang, Zufeng Zhang, Ping Zhou, Haowen Jiang Key Laboratory of Metallurgical Equipment and Control Technology, Wuhan University of Tianjin University of Technology,Binshui Xidao391, Tianjin, China Science and Technology, Wuhan, China • The intelligent air purifier is based on STM32F407 MCU. • Establish the mathematical model of • After tested, the air condition is path planning of tube-sheet welding. divided into three grades: superior, • Introduce the standard firefly good and bad. algorithm. • The user can choose the automatic • Improved the firefly algorithm to the mode or manual mode to adjust the discrete firefly algorithm. purification level.The test results will be transmitted to the mobile • The experiment of path planning of Intelligent Air Purifier System terminal. tube-sheet welding is taken by applying of the proposed DFA. The tube-sheet welding equipment • The results show that the effective purification rate of the air purifier is 97%, and the sterilization rate is 99%.

FA1-3(5) 9:30-9:45 FA1-3(6) 9:45-10:00 Simulation and Experimental Research on the Fatigue Life of Hydraulic Excavator Undercarriage Thread Turning of Planetary Roller Screw Yao Li , Bing Ma, Zhengbing Xia, Jianjian Wang, Yufei Hao, Jinfei Cao, Zongxin Yu, Xueyong Yi, Kun Zeng Mechanism Key Lab. of Road Construction Technology and Equipment, Chang’an University Xi’an, China Xiaoshuai Duan1, Xinhua Zhang1, Minghui Hao2, Xijian Huo1 and Xiaobin1 1.Beijing Institute of Automation and Control Equipment, Beijing, China • The fatigue bench test of undercarriage was 2.Harbin Institute of Technology, Harbin, China designed by reverse loading, which makes it • The unit cutter model and the main convenient to apply fatigue test spectrum to cutting force model of thread. undercarriage. • The finite element simulations • Using data processing methods such as rain- calculation of thread turning. flow counting, combined with the damage equivalence and damage consistency • The turning experiments of the calibration principle, the sampled data were screw rod. compiled into constant amplitude spectrum • The finite element simulations can for the undercarriage. FEM of a undercarriage effectively ensure thread processing Fig.1 Tool tip wear at different • The fatigue life of the undercarriage was process PRSM. cutting depths predicted.

IEEE ICMA 2020 Conference Digest FA1-4: Humanoid Robots

Session Chairs: Jian Guo, Tianjin University of Technology Qiang Huang, Beijing Institute of Technology Online Conference Room 4, UTC+8(Beijing Time): 8:30-10:00, Friday, 16 October 2020

FA1-4(1) 8:30-8:45 FA1-4(2) 8:45-9:00 Desire-Driven Reasoning considering Status- Motion Sequence Learning for Robot Walking Based Knowledge Description for Personal Care Based on Pose Optimization Robots Yancao Jiang1, Weiyi Zhang1, Fasih Ud Din Farrukh1, Xiang Xie1 and Chun Zhang1 Institute of Microelectronics，Tsinghua University Guang Yang*, Shuoyu Wang*, Junyou Yang** and Peng Shi*** Beijing, China *Kochi University of Technology, Kami, Japan **Shenyang University of Technology, Shenyang, China ***The University of Adelaide, Adelaide, Australia • Bipedal robot • Reinforcement learning • Developing personal care robots is • neural network one option to face the aging society • pose optimization and caring force. • A desire-driven reasoning system considering status-based knowledge description is proposed. • The reasoning system was evaluated The Personal Care Robot The Robotis-op3 Robot with personal care robot KUT-PCR in a real household domain.

FA1-4(3) 9:00-9:15 FA1-4(4) 9:15-9:30

Design of Graphical User Interface for Motor Selection of the Lower Limb Design and Analysis of a Lower Limb Exoskeleton Exoskeleton Robot Jian Guo1, Fan Bu1, Shuxiang Guo1,2 Shuxiang Guo1,2*， Yibin Ding1 , Jian Guo1 1Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Intelligent Robot Laboratory ， 1 Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Intelligent Robot Laboratory, Tianjin University of Technology Tianjin, China 2 Tianjin University of Technology, Tianjin, China Department of Intelligent Mechanical Systems Engineering , Faculty of Engineering 2 Department of Intelligent Mechanical Systems Engineering, Faculty of Engineering, Kagawa University , Japan Kagawa University, Kagawa, Japan • This paper proposes a wearable • For the purpose of determine the motor type, this lower limb exoskeleton robot. paper developed a MATLAB Graphical User Interface(GUI). • Adams and Solidworks software are • Kinematics analysis is to analyze the relationship used to perform static stress analysis between the rotation angle of the rotary joint and and gait walking simulation of the the position of the connecting rod in this GUI. structure. • Dynamics enables the control robot to follow the required trajectory more accurately. • The motion parameters of the joints • The motion of each joint can be represented by a are obtained through simulation, polynomial by Taylor functions. Therefore, only which provides a basis for the the coefficients of the polynomial need to be Data flow diagram. input, and the polynomial of the motor torque can design of the structure. be obtained through calculation. The 3D model of the robot

FA1-4(5) 9:30-9:45 FA1-4(6) 9:45-10:00 A Novel Foot Contact Probability Estimator The study of impedance control of lower for Biped Robot State Estimation extremity exoskeleton Yali Han, Jinfei Shi, Han Sun, Weijie Zhou, Hongyao Guan, Chuanqi Shi Mingyue Qin, Zhangguo Yu, Xuechao Chen, Qiang Huang, Chenglong Fu, School of mechanical engineering,Nanjing institute of technology Aiguo Ming and Chunjing Tao Nanjing, China School of Mechatronical Engineering, Beijing Institute of Technology, Beijing, China • A lower limb rehabilitation exoskeleton based on motor and cable drive was  • Propose the flat foot constrained right B designed. invariant Kalman filter. • An impedance control algorithm based on pt • Propose a simplified foot contact z robust adaptive compensation was designed. probability estimator based on the y • Establishing a joint simulation platform friction and center of mass constraint.  G based on virtual prototypes of Matlab and z y • Optimize the parameters of the x t t d Adams. t  estimator through minimizing the F x • The experiment result of the exoskeleton floating base state estimation error by t showed that hip exoskeleton can follow the Bayesian optimization. Frame and State Defination human angle well. of Flat Foot Bepied Robot Lower extremity exoskeleton

IEEE ICMA 2020 Conference Digest FA1-5: Rotor Dynamics, Vibration Analysis and Vibration Control

Session Chairs: Wei Zhou, Beijing Institute of Technology Yan Zhao, Beijing Institute of Technology Online Conference Room 5, UTC+8(Beijing Time): 8:30-10:00, Friday, 16 October 2020

FA1-5(1) 8:30-8:45 FA1-5(2) 8:45-9:00 Research on Asymmetrical Dual-rotor System Research on Crack Detection of Nonlinear Rotor Based on the Energy Orbit Based on Energy Tracks Zhu Han1,2, Jun Liu1,2*, Yu Zhang1,2 Yu Zhang1,2, Jun Liu1,2*, Zhu Han1,2 1. Tianjin Key Laboratory for Advanced Mechatronic System Design and Intelligent Control, 1. Tianjin Key Laboratory for Advanced Mechatronic System Design and Intelligent Control, Tianjin University of Technology, Tianjin, 300384, China. Tianjin University of Technology, Tianjin, 300384, China. 2. National Demonstration Centre for Experimental Mechanical and Electrical Engineering 2. National Demonstration Centre for Experimental Mechanical and Electrical Engineering Education, Tianjin University of technology, Tianjin, China. Education, Tianjin University of technology, Tianjin, China.

• An novel asymmetrical dual-rotor • The concept of energy track was 0.5 Dk=0.05 coupling dynamic model is proposed. introduced to the nonlinear rotor Dk=0.1 0.0 Dk=0.15 • The vibration responses of the system system with crack.

are analyzed systematically. • The vibration characteristics with -0.5 Displacement(y) • The concept of energy orbit was first different crack parameters in are -1.0

introduced into the dual-rotor system. compared based on energy tracks. -1.5 -1.0 -0.5 0.0 0.5 1.0 Displacement(x) • The vibration characteristics of system • The vibration characteristics in are analyzed and investigated. based different critical speeds are Vibration Energy Tracks on the energy orbit. Vibration Energy Orbit compared based on energy tracks. With Crack

FA1-5(3) 9:00-9:15 FA1-5(4) 9:15-9:30 Bearing Selection of 20KW High Speed Motor Dynamic Simulation of Tooth Root Crack Failure and Analysis of Rotor Dynamic Characteristics in Planetary Gearbox Based on Torsional Jiayi Yao, Haipeng Geng and Wei Zheng New Energy Equipment and Quality Engineering Research Institute, Xi’an Jiaotong University Vibration Signal Xi’an, China Jingyuan Liang, Yunpeng Li(2), Jianchuan Dai, Hang Niu and Chenggang Hou • Bearing should be selected Department of Mechanical engineering, Xian Jiaotong University (1) China Academy of Machinery Science and Technology Group Co., LTD (2) reasonably according to the Xi’an, China, Qingdao, China working load, working speed, etc. • Rigid-flexible coupling model. • The high running speed • Torsional vibration analysis. makes the motor more and • Diagnose tooth root crack failure in more demanding on the planetary gearbox. rotor dynamic characteristics. • In order to reduce the design margin, the first six natural frequencies The Model of Planetary Gearbox are calculated under different bearing steel stiffness, providing a reference for bearing stiffness selection.

FA1-5(5) 9:30-9:45 FA1-5(6) 9:45-10:00 Research on Rotor Dynamics of a 60000rpm Analysis and Simulation Verification of 200KW High Speed Motor Mechanical Properties for Planetary Roller Dingchong Lyu, Haipeng Geng, Qiyuan Liu Screw Mechanism School of Mechanical Engineering, Xi'an Jiaotong University Xi'an, Shanxi Province, China Xiaoshuai Duan1, Xinhua Zhang1, Minghui Hao2, Xijian Huo1 and Ping Guan1 1.Beijing Institute of Automation and Control Equipment, Beijing, China • Simplified finite element model of 2.Harbin Institute of Technology, Harbin, China the rotor is constructed; • Distribution characteristics of PRSM. • The modal analysis is carried out, the rotor system is relatively stable, • Influence of pitch and roller thread the critical velocities are solved; length on the load capacity of PRSM. • Influence of bearing stiffness and • The stress distribution of the thread damping on the rotor dynamic contact point is improved, and the characteristics is analyzed; load capacity is increased to 1.2 times of that before modification.. • Harmonic response analysis is Model section of the motor Fig.1 Equivalent stress nephogram conducted. on the side of roller and nut

IEEE ICMA 2020 Conference Digest FA2-1: Neuro, Fuzzy, and Intelligent Control (II)

Session Chairs: Shuoyu Wang, Kochi university Shuoxin Gu, Chengdu University of Information and Technology Online Conference Room 1, UTC+8(Beijing Time): 10:30-12:00, Friday, 16 October 2020

FA2-1(1) 10:30-10:45 FA2-1(2) 10:45-11:00 Formating Reshape Control Strategy for Multiple Brushless DC motor Control System Based on Amphibious Spherical Robots RBF Neural Network Zan Li, Shuxiang Guo, Liwei Shi Key Laboratory of Convergence Medical Engineering System and Healthcare Technology, Hong Xu, Linsen Xu*, Shouqi Chen, Jinfu Liu , Gaoxin Cheng, Xingcan Liang, and Wujun Mao Beijing Institute of Technology University of Science and Technology of China Beijing, China Hefei, China • This paper puts forward a new strategy that a formation of • Spherical joints with two degrees of multiple amphibious spherical freedom were designed. robots pass through a narrow • The RBF neural network self-tuning waterway. control system is designed for • Virtual linkage structure and Virtual Linkage Structure Scenario BLDC motor. consensus algorithm are used to • Through simulation analysis, the plan and optimize the trajectory self-tuning PID control and PID of whole formation. control are compared to analyze the • Auto disturbance rejection performance difference between Spherical joints with two controller is used in each robot them. degrees of freedom to track its own trajectory. Formating Control Strategy

FA2-1(3) 11:00-11:15 FA2-1(4) 11:15-11:30

Study on Force Feedback Control of the Vascular Study on Delay and Accuracy of Control System Interventional Surgical Robot based on Fuzzy PID for the Vascular Interventional Surgical Robot based on Fuzzy 1 1 1,2* PID and Improved Smith Algorithms Jian Guo ,Lixin He ,Shuxiang Guo Jian Guo1, Zixiang Zhuang1 , Shuxiang Guo1,2* 1.Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and 1.Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Intelligent Robot Laboratory Tianjin University of Technology Intelligent Robot Laboratory ,Tianjin University of Technology 2. Department of Intelligent Mechanical Systems Engineering, Faculty of Engineering Tianjin, China Kagawa University • Improved Smith algorithm was proposed to 2.Department of Intelligent Mechanical Systems Engineering, Faculty of Engineering, Kagawa e University improve the real-time of control system for time delay phenomenon which caused by motor of the d/dt Ec Fuzzy PID Kagawa, Japan control slave side and also the data processing from the vascular interventional surgical robot kp ki kd Motor transfer • Fuzzy PID algorithm was proposed to improve function R(S) + + ⚫ This paper purposed that improved the accuracy of force the accuracy of control system for vascular Y(S) C(S) G() s e− s feedback and enhanced the safety of surgery. interventional surgical robot. The fuzzy PID and PID applied to force feedback control Control ⚫ • Delay and accuracy analysis was carried out by - - node + are compared with dynamic simulation and the comparison using algorithms and experimental method, and − ms results are analyzed. Gm () s e the experimental results showed that the system - ⚫ Fuzzy PID and PID control are respectively applied to the with Fuzzy PID and improved Smith algorithms Gm(S) actual force feedback control, and the experimental results improved the real-time and accuracy for the

are analyzed. vascular interventional surgical robot. Fig.Control system of vascular interventional surgical robot based on fuzzy PID and improved smith algorithms Simulation comparison model of controller

FA2-1(5) 11:30-11:45 FA2-1(6) 11:45-12:00 Continuity and Smoothness Analysis and Short-term Prediction of Photovoltaic Power Possible Improvement of Traditional Based on PLSR Improved DBN Reinforcement Learning Methods Wu Sun, Jianwei Leng Start DBN parameter School of Electrical And Electronic Engineering initialization Tianhao Chen, Wenchuan Jia, Jianjun Yuan, Shugen Ma, Limei Cheng Tianjin University of Technology Update the RBM School of Mechatronic Engineering and Automation connection weight for Shanghai Key Laboratory of Intelligent Manufacturing and Robotics • This paper proposes a partial least each layer The PLSR model is Shanghai University，Shanghai, China squares regression method instead of established between RBM • Explain from mathematical theory BP method to perform parameter and output in the last layer Bottom - up parameter why RL methods for stochastic tuning method and Cross-verify its tuning based on PLSR strategies generally do not have rationality. Cross validation continuity conditions. Y The optimal DBN • The improved fuzzy C-means parameters were obtained • Discuss the smoothness of stochastic and the model was trained clustering (FCM) algorithm was policies of these main RL methods DBN-PLSR after used to cluster the input sample data training based on time-frequency diagrams of Example prediction and results action sequences. then the Photovoltaic power prediction model of FCM-DBN- end • Propose a new RL scheme that can solve the problems of continuity and PLSR was established. Prediction process Morlet scalogram of action smoothness possibly. structure diagram.

IEEE ICMA 2020 Conference Digest FA2-2: Soft Robotics

Session Chairs: Aiguo Ming, University of Electro-Communications Peng Shi, Kagawa University Online Conference Room 2, UTC+8(Beijing Time): 10:30-12:00, Friday, 16 October 2020

FA2-2(1) 10:30-10:45 FA2-2(2) 10:45-11:00

An untethered bionic robotic fish Kinematics Solution of Snake-like Manipulator using SMA actuators Based on Improved Backbone Mode Method Xiaojie Chen, Hiroki Shigemune, Hideyuki Sawada Zhen Wang, Jian Chang, Bin Li, Cong Wang and Chun Liu Department of Applied Physics, Waseda University, Tokyo, Japan State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Shenyang, China

• Fabrication of a silicone-made tail Inside Battery and Controller fin using shape-memory alloy wires. • Snake-like manipulator has the • Development of an untethered bionic advantage of high degrees of robotic fish and its wireless control. freedom, suitable for some complex • Two underwater experiments to working environment. examine the factors affecting the • the kinematic model of snake-like propulsion ability. manipulator • The robotic fish realized the BCF • The improved backbone mode swimming mode, which was method to solve the kinematics commonly observed in nature fish. Outer view of the robotic fish and its swimming motion The snake-like manipulator.

FA2-2(3) 11:00-11:15 FA2-2(4) 11:15-11:30

Study on the Active Movement Capsule Robot Design of a Novel Micro Robot In-pipe for Biopsy Shuxiang Guo1,2, Yaqi Hu1, Jian Guo1, Qiang Fu1 Shuxiang Guo1,2, Fang Huang1, Jian Guo1, Qiang Fu1 1 Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Intelligent Robot Laboratory, 1 Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Intelligent Robot Laboratory, Tianjin University of Technology, Tianjin, China Tianjin University of Technology, Tianjin, China 2 Department of Intelligent Mechanical Systems Engineering, Faculty of Engineering, 2 Department of Intelligent Mechanical Systems Engineering, Faculty of Engineering, Kagawa University, Kagawa, Japan Kagawa University, Kagawa, Japan • A active movement capsule robot for • A novel micro in-pipe robot with a biopsy diven by an external magnetic propeller is proposed in this paper, it is a field is proposed. robot that can actively move. • The external electromagnetic control • Fluid simulation model is established. system and the structure and movement • The feasibility of the capsule robot is principle of biopsy capsule robot were verified through experiments. analyzed. Schematic diagram of the • The evaluation experiment of robot micro-robot Schematic diagram of accurate biopsy were completed. biopsy robot

FA2-2(5) 11:30-11:45 FA2-2(6) 11:45-12:00 A Replaceable Vascular Model-based Platform for Behavior Analysis of Soft Pneumatic Actuator Experience Acquisition in Interventional Surgery Gripper by using Image Processing Technology Yue Wang1, Jin Guo*, Shuxiang Guo1,2, Chuqiao Lyu1, Youchun Ma1 Hoai Nam Le, Quoc Van Vo, Hoang Phuc Van Ngo, Nhu Thanh Vo, Anh Duc Pham Key Laboratory of Convergence Biomedical Engineering System and Healthcare Technology, The University of Danang – University of Science and Technology The Ministry of Industry and Information Technology, School of Life Science, Danang, Vietnam Beijing Institute of Technology, Beijing, China • We collected surgical experience • This study presents a systematic approach to analyze the behaviors of from the perspective of interventional a SPA gripper used in the food surgical image information. industry by using image processing. • In this paper, an experience • Firstly, the structural geometry of acquisition platform was established, soft gripper model controlled by on which surgeons could observe pneumatic actuators is suggested. surgical images to facilitate • Moreover, the geometrical postoperative learning. parameters like chamber thickness • Surgeons can use the platform to of the soft gripper is optimized by GUI of Experience Acquisition Platform the combination of finite element mark the movement of guidewire so Diagram of analysis process method (FEM) modeling and as to evaluate the operation of experiment. surgeons.

IEEE ICMA 2020 Conference Digest FA2-3: Sensor Networks, Distributed Sensor Systems

Session Chairs: Takayoshi Yokota, Tottori University Songyuan Zhang, Harbin Engineering University Online Conference Room 3, UTC+8(Beijing Time): 10:30-12:00, Friday, 16 October 2020

FA2-3(1) 10:30-10:45 FA2-3(2) 10:45-11:00 Localization Algorithm Based on Altitude Profile Indoor Pedestrian Positioning Method Based on Matching Derived from Barometer Pelvic Motion Model Using Inertial Sensors Takayoshi Yokota, Miao Yang1,2, Lelai Zhou1,2 and Yibin Li1,2 Department of Information and Electronics, Graduate School of Engineering, 1.School of Control Science and Engineering, Shandong University Tottori University, Tottori, Japan 2.Engineering Research Center of Intelligent Unmanned System, Ministry of Education Jinan, Shandong Province, China • Localization algorithm based on • The center of the pelvic is selected as the altitude matching is proposed. root point. • Altitude is obtained with a • The five sensors are located at the shank, barometer(atmospheric pressure thigh and the pelvis to measure the sensor). kinematics of the lower limbs. • The position estimation error is Concept of altitude profile matching • The roll obtained from the sensor is used about 20m in the area where the to estimate the step length and the yaw is altitude change is remarkable. used to estimate the heading of pedestrian • Dynamic Programming based movement. The Motion Model profile matching algorithm in • The position is determined by step length spatial domain is applied. information and heading information. Evaluation Results (Position Error)

FA2-3(3) 11:00-11:15 FA2-3(4) 11:15-11:30 Study on Distributed Data Processing System for Decentralized Spherical Multi-robot based on Edge Coverage Optimization Algorithm for Multi-robot Computing and Blockchain System based on Virtual Force Refinement Shuxiang Guo1,2 ,Sheng Cao1 ,Jian Guo1* ,Jigang Xu3* Tonghui Zeng, Xiaomei Xie*, Mingzhu Wei, Xin Chen, Xiangfei Wu 1Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and School of Aeronautics and Astronautics of UESTC Biomedical Robot Laboratory, Tianjin University of Technology, Tianjin, China Chengdu, China 2Intelligent Mechanical Systems Engineering Department Faculty of Engineering, Kagawa University, Kagawa, Japan • Initial coverage of Lloyd-Voronoi 3Qinghai Haidong,810700,China algorithm. • This paper proposed the distributed data • Determination of the minimum processing system of the spherical multi-robot based on Edge Computing number of robots by embedding and Blockchain. polygons in the circle perception • Network architecture of spherical multi- model robot system based on blockchain can avoid the possibility of invasion, solve • Virtual force algorithm to repair the Byzantine general problem and improve the flexibility and intelligence. the coverage hole • The results show that distributed data processing system can effectively Coverage optimization algorithm improve the real-time performance. The architecture of edge computing

FA2-3(5) 11:30-11:45 FA2-3(6) 11:45-12:00 A Method of Lower Limb Gait Based on Multi- A Multi-model Human Motion Tracking Approach sensor Data Fusion for Rehabilitation Robot with Wearable IMU Sensors 1 1 1 1 1,2 Jiale Lv , Jian Li , Shigang Wang , Rentao Wang , Xueshan Gao Xue Li, Lelai Zhou,Yibin Li 1 Robot Institute of Guangxi University of science and technology 1School of Control Science and Engineering, Shandong University Liuzhou,Chain 2Engineering Research Centre of Intelligent Unmanned System, Ministry of Education, 250061 2 School of Mechatronical Engineering of Beijing Institute of Technology Jinan, Shandong Province, China Beijing,Chain • Using integrated gyroscope, accelerometer • Four kinematic models are proposed. and magnetometer as sensors, the • Coordinate transformation is used to installation position is shown in the yellow track the attitude changes of human circle. body according to the attitude of IMU. • The four models are switched • Ellipsoid fitting based on least square dynamically and the positions are method is used for calibration, quaternion obtained according to human kinematics and the quadratic integral of method is used for attitude calculation, and and acceleration. second-order Kalman filter is used for • The model switching mechanism is Human motion tracking attitude compensation. It can collect determined according to the change of characteristic states such as turning around foot angular velocity. and walking unstable during walking Gait information collection

IEEE ICMA 2020 Conference Digest FA2-4: Human-System Interaction and Interface

Session Chairs: Yong Yu, Kagoshima University Cheng Yang, Beijing Institute of Technology Online Conference Room 4, UTC+8(Beijing Time): 10:30-12:00, Friday, 16 October 2020

FA2-4(1) 10:30-10:45 FA2-4(2) 10:45-11:00 Speech Analysis of the Talking Robot with Gesture Recognition based on BP Neural Human-like Artificial Vocal Tract Network and Data Glove Qiang Fu, Jiajun Fu , Jian Guo, Shuxiang Guo , Xun Li 1 2 Thanh Vo Nhu and Hideyuki Sawada Tianjin Key Laboratory for Control Theory & Application in Complicated Systems 1. The University of Danang, Vietnam, 2. Waseda University, Japan and Biomedical Robot Laboratory • A mechatronic vocalization system Tianjin University of Technology Tianjin China has been developed by referring to the human speech mechanism. Various devices for hand data collection have been proposed for gesture • An artificial vocal tract having the recognition. In this paper, a method is same shape and size of an adult was proposed for gesture recognition based on made for this robotic speaking system. data glove and BP neural network. This • A mapping of sounds corresponding data glove can collect data of the hand and to robotic parameters was created for forearm, it gives a comprehensive automatic vocalization. Human-like Artificial Vocal Tract description of all aspects of arm • Formant frequencies of human and for Speech Generation movement, and BP neural network robotic sounds were extracted for the algorithm is used to process and classify comparison and evaluation. the collected data and compress the data. The Data glove

FA2-4(3) 11:00-11:15 FA2-4(4) 11:15-11:30 An Improved Visual Auxiliary Algorithm for the Vascular A Robust Multi-Channel EEG Signals Interventional Surgical Robot based on Neural Network Preprocessing Method for Enhanced Upper Jian Guo1, Suxiang Feng1 Shuxiang Guo1,2 1Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Intelligent Robot Laboratory Extremity Motor Imagery Decoding 1,2,3 1,2 1,2 1,2 Tianjin University of Technology ，Tianjin, China Mojisola Grace Asogbon ,Oluwarotimi Williams Samuel , Xiangxin Li ,Yanjuan Geng , 2 Department of Intelligent Mechanical Systems Engineering Oluwagbenga Paul Idowu1,2,3,Yanbing Jiang1,2,3,Naifu Jiang1,2, Ali H. Al-Timemy4 &Guanglin Li1,2* Kagawa University Japan 1Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences (CAS), China. • This paper proposes to apply the 2Key Laboratory of Human-Machine Intelligence-Synergy Systems, CAS, Shenzhen, China 3 neural network algorithm into the u Shenzhen College of Advanced Technology, UCAS, Shenzhen 518055, China o x 4 visual auxiliary coordinate p Biomedical Engineering Department, University of Baghdad, Baghdad 47146, Iraq. f calibration of the vascular ATSC ACA = 77.56% v y • A robust method, GEVD-MWF is proposed Oc Xc interventional surgical robot. for multiple artifact removal from EEG 100 Yc 80 • Making mathematical modeling Zc signals, for accurate MI task decoding. for the visual auxiliary system. 60 • A modified channel selection technique, 40

• Comparing the deviation between 20 mSFFS was implemented for optimal EEG P the actual coordinates collected by Yw 0 NDI and the algorithm fitting channel selection. Xw coordinates. Ow Zw • The results showed that the proposed method ASC ACA = 99.88% RCASC ACA = 96.87% • The experimental results show that The coordinate systems of vascular led to substantial increase in performance the visual auxiliary based on interventional robot assist system compared to existing approach. With the Average Classification neural network algorithm is mSFFS, an accuracy of 96.87% was Accuracy (ACA) for ATSC, reliable. achieved with 10 optimal channels. ASC and RCASC.

FA2-4(5) 11:30-11:45 FA2-4(6) 11:45-12:00 Finite Element Analysis of Series Elastic Actuator for Exoskeleton Robot Joint Calibration Method of Binocular Fisheye Lens Pengyu Xi, Weimin Ge, Sen Zhang, Zhili Zhang, Teresa Zielinska Based on Global Parameters Tianjin Key Laboratory for Advanced Mechatronic System Design and Intelligent Control, Xiaoming Xu, Jinyu Zhang,Xuemei Wang Tianjin University of Technology, Tianjin, China Institute of Physical and Chemical Engineering of Nuclear Industry • Use Opensim software to simulate Tianjin, China the gait of the human body to obtain the torque of each joint of the human • A whole parameter calibration body when the lower limbs are method is adopted to calibrate the walking normally binocular fisheye lens. • Use the human joint torque obtained • The binocular fisheye lens is by Opensim as the input torque of modeled. the exoskeleton robot joint • The parameters of binocular fisheye component, and perform finite lens are obtained, which solves the element analysis on the component problem that the conventional The experimental platform Stress strain diagram calibration method cannot be directly applied to the fisheye lens.

IEEE ICMA 2020 Conference Digest FA2-5: Control Theory and Application (V)

Session Chairs: Lingling Zheng, Kagawa University Xianqiang Bao, Beijing Institute of Technology Online Conference Room 5, UTC+8(Beijing Time): 10:30-11:30, Friday, 16 October 2020

FA2-5(1) 10:30-10:45 FA2-5(2) 10:45-11:00

Fault location based on energy relative entropy Image Segmentation Method of Crop Diseases Hui Zhao, Lei Wang, Hongjun Wang,Youjun Yue Based on Improved Segnet Neural Network Tianjin University of Technology, Youjun Yue,Xuesong Li, Hui Zhao, Hongjun Wang, Tianjin Key Laboratory of Complex System Control Theory and Application, Tianjin University of Tianjin, China Input training Technology Tianjin, China image Use the segnet model to encode the volume base layer and • Use the PMU to collect the transient pooling layer in the network to extract features and save (1) Based on segnet network, this paper the pooling index zero sequence current after the Perform deconvolution and upsampling in the combines segnet with conditional segnet model decoding network to obtain the faulty line fails. segmentation map. random field to improve the network to Input the result of pixel classification into CRF • Use the EMD decomposition segment the crop disease image. Calculate CRF first-order function and second-order function with algorithm to decompose the (2) Among all the segmentation pixel classification results uploaded zero sequence current. Constrain the pixel classification methods, the three index values P, R, results with a first-order function and a second-order function

• According to the relative entropy and F of the improved segnet network Combine segnet and CRF models to form an improved segnet network model structure theory, calculate the energy relative Zero sequence current are the highest among all methods. for training and complete optimization Enter Optimize the model parameters entropy of the adjacent node signal. (3) The method in this paper corrects verification of segnet and CRF through the decomposition image verification process

Use the improved segnet the shortcomings of poor adaptability Input test • Determine the section with the network model after parameter image in the general method, and has good tuning to segment the test image largest relative energy entropy as the Output test image fault section. robustness. segmentation results Flow chart of this method

FA2-5(3) 11:00-11:15 FA2-5(4) 11:15-11:30 Trajectory-tracking and Coordinated Control of Technology Multi-channel precision Six-wheel-drive Rocker-bogie Patrol Robot temperature control system based on TEC Weidi Chen, Song Jiang, Zhenghui Shen, Xuejiao Yan and Meng Chen Aerospace System Engineering Shanghai, Shanghai Academy of Spaceflight Technology Xiaoming Xu, Jinyu Zhang,Xuemei Wang Shanghai, China Institute of Physical and Chemical Engineering of Nuclear Industry Tianjin, China • A six-wheel-drive rocker-bogie patrol robot system is designed. • Fuzzy algorithm is used to • Hybrid PD control and LQR optimal automatically control and adjust PID control strategies are used to achieve parameters high-precision trajectory tracking . • Experiments show that the • A coordinated controller is designed temperature stability of the system is to realize the speed distributing of six better than plus or minus 0.002 driving wheels. degrees in the temperature control • Simulation results show that it can The Patrol Robot range of-20 ~ 120 degrees celsius. multi channel precise temperature controller obtain high trajectory-tracking accuracy and effectively reduce the power loss of multi-drive system.

Index of Session Chairs Guo, Jian FA1-4 -A- Guo, Jin WP1-1 An, Ruochen WP2-5 Guo, Jin FA1-1 An, Ruochen TP1-3 Guo, Shuxiang WP3-1 Guo, Shuxiang TA2-5 -B- Bao, Xianqiang TP2-5 -H- Bao, Xianqiang FA2-5 Hirata, Hideyuki WP1-5 Bu, Dongdong WP3-5 Hirata, Hideyuki TP2-4 Bu, Dongdong TA2-6 Hou, Xihuan WP1-6 Hou, Xihuan TP1-4 -C- Huang, Jian TP2-1

Cao, Liwen WP1-3 Huang, Qiang TP2-6 Huang, Qiang FA1-4 -D- -I- Dong, Enzeng TP2-2 Imai, Shinichi TA2-2 -F- -J- Fu, Qiang WP2-4 Jia, Chao WP3-3 Fu, Qiang FA1-3 Jia, Chao TP1-1 Fukuda, Toshio WP1-2 Jin, Xiaoliang WA1-P Fukuda, Toshio TA1-1 Jin, Xiaoliang TP1-6 Fu, Yili WP1-6 Jin, Xiaoliang FA1-2 Fu, Yili TA1-4

-G- -K- Kosuge, Kazuhiro WP2-1 Gao, Xueshan TA2-6 Kosuge, Kazuhiro TA1-6 Gao, Xueshan TA2-3 Gold, Tobias WP2-2 -L- Gu, Shuoxin TA2-3 Li, Jian WP1-3 Gu, Shuoxin FA2-1 Li, Jian TA2-1 Guo, Jian TA1-5 Li, Qin WP1-1

Li, Qin TA1-2 Wang, Shuoyu TP1-2 Lin, Hsien-I WP1-4 Wang, Shuoyu FA2-1 Liu, Yi WP3-2 Wang, Zixu WP2-3 Liu, Yi TA1-6 Wang, Zixu TP2-4 Luo, Dingsheng TA1-1 Watanabe, Keigo WP1-4 Lyu, Chuqiao TP2-2 Watanabe, Keigo TA1-3 Lyu, Chuqiao TP1-5 Wu, Haiyuan WP2-4 Wu, Qiong TA1-5 -M- Wu, Qiong TA2-5 Mills, James K. WP3-5 Mills, James K. TP1-5 -X- Ming, Aiguo TP1-3 Xing, Huiming WP3-1 Ming, Aiguo FA2-2 Xing, Huiming TA2-4 -N- -Y- Nagata, Fusaomi WP2-6 Yang, Cheng WP2-1 Yang, Cheng FA2-4 -P- Ye, Xiufen WP2-5 Pan, Qinxue WP2-3 Ye, Xiufen TP2-3 Pan, Qinxue TP1-1 Yin, He TA1-2 Yin, He FA1-3 -S- Yokota, Takayoshi FA2-3 Sawada, Hideyuki WP3-6 Yu, Qingguang TA2-4 Sawada, Hideyuki TP1-6 Yu, Qingguang WP3-4 Shi, Chaoyang FA1-2 Yu, Yong WP2-6 Shi, Chaoyang WP1-2 Yu, Yong FA2-4 Shi, Liwei WA1-P Shi, Liwei TP1-4 -Z- Shi, Peng TA2-1 Zhang, Lanyong TP1-2 Shi, Peng FA2-2 Zhang, Lanyong TA1-3 Song, Yu TA2-2 Zhang, Linshuai TP2-5 Song, Yu WP3-3 Zhang, Songyuan TP2-1 Zhang, Songyuan FA2-3 -W- Zhao, Yan WP3-2 Wang, Hongjun FA1-1 Zhao, Yan FA1-5

Zheng, Liang TP2-3 Zheng, Liang WP3-6 Zheng, Lingling WP3-4 Zheng, Lingling FA2-5 Zhou, Junjie WP2-2 Zhou, Junjie TA1-4 Zhou, Wei TP2-6 Zhou, Wei FA1-5 Zhou, Wei WP1-5

Index of Authors

-A- -C- Æsøy, Vilmar TA1-1 Cai, Huimin WP1-1 Ali, Haider WA1-P Cai, Liming TA2-6 Al-Timemy, Ali H. FA2-4 Cai, Shiyuan TP2-2 An, Ruochen WP3-6 Cai, Zhuocong WP1-1 An, Ruochen TA1-3 Cai, Zhuocong TP2-4 An, Ruochen TP1-3 Cao, Fengqiu WA1-P An, Zijing WP2-1 Cao, Huifeng TP2-3 Arai, Tatsuo FA1-2 Cao, Jinfei FA1-3 Arumalla, Kirit WP3-2 Cao, Liwen WA1-P Asogbon, Mojisola Grace FA2-4 Cao, Sheng FA2-3 Awa, Tendeng TP1-3 Cao, Zhiqiang WP3-1 -B- Cao, Zhiqiang WP3-4 Chang, Jian WP1-3 Bai, Dan WP3-6 Chang, Jian FA2-2 Bai, Jing TP1-2 Chang, Meile WP3-3 Bai, Xinyu TA2-3 Che, Hongjuan WA1-P Bao, Jieqiu WA1-P Che, Hongjuan WP1-2 Bao, Jieqiu WA1-P Che, Hongjuan TP1-5 Bao, Jieqiu FA1-1 Chen, Chao TP1-3 Bao, Weijie FA1-3 Chen, Chao TA2-4 Bao, Yan WA1-P Chen, Chenfei TP1-5 Bao, Yan FA1-1 Chen, Daye WP3-2 Bi, Junhao TA1-5 Chen, Feng TP1-3 Bi, Yongli WP2-5 Chen, Guangsong TA2-4 Bian, Guibin WP1-1 Chen, Guoguang TP1-5 Biglarbegian, Mohammad WA1-P Chen, Hao WP3-4 Bolchover, Paul TA1-1 Chen, Huajun WP1-3 Bu, Dongdong WA1-P Chen, Huiling WA1-P Bu, Dongdong TA2-6 Chen, Huiling TA1-3 Bu, Fan FA1-4 Chen, Jiao TA2-3 Busch, Alexander WA1-P Chen, Jingwei TP2-4

Chen, Junhua TA2-4 Cheng, Gaoxin FA2-1 Chen, Jyh-Cheng WP2-3 Cheng, Handong WA1-P Chen, Kailun WP1-5 Cheng, Limei FA2-1 Chen, Li TP2-1 Cheng, Long WP3-4 Chen, Longmiao TA2-4 Cheng, Yaozong TP1-1 Chen, Meng FA2-5 Cheng, Yutao WP2-2 Chen, Peng WA1-P Chiang, Cheng-Ta WP2-3 Chen, Peng WA1-P Chu, Lingling TP2-1 Chen, Qian WP2-4 Chu, Yuyi TA2-6 Chen, Rongbing TA1-1 Chung, Won Jee WP3-5 Chen, Shouqi WA1-P Cui, Dekang TA1-4 Chen, Shouqi WP3-5 Cui, Hang WP2-1 Chen, Shouqi WP1-6 Cui, Liang FA1-2 Chen, Shouqi FA2-1 Cui, Mingyu TA2-2 Chen, Tianhao FA2-1 Cui, Mingyu TP1-3 Chen, Weidi FA2-5 Cui, Nianqiang TP1-3 Chen, Xi TP1-1 Cui, Song WP2-1 Chen, Xi TP1-5 Cui, Weimin WP1-2 Chen, Xiangyu TP2-3 -D- Chen, Xiaojie FA2-2 Chen, Xin WP2-4 Dai, Jianchuan FA1-5 Chen, Xin TA2-3 Dai, Ruizhen TP2-4 Chen, Xin FA2-3 Dai, Xianze WA1-P Chen, Xinnian TP1-4 Dai, Xiaobo WA1-P Chen, Xuechao WP3-1 Dai, Yuehong WP1-5 Chen, Xuechao FA1-4 Dai, Yuehong FA1-1 Chen, Yipeng WP2-4 Dang, Yanyang FA1-1 Chen, Yulong TA2-1 Deng, Hanbo WP3-6 Chen, Zhengyang TP1-6 Deng, Yian TA1-1 Chen, Zhengyang TP2-6 Ding, Guanqi TP1-2 Chen, Zhilei TA1-3 Ding, Min WP3-4 Chen, Zhong WP3-2 Ding, Yibin FA1-4 Cheng, Gaoxin WA1-P Dong, Chao WA1-P Cheng, Gaoxin WP3-5 Dong, Chao WA1-P Cheng, Gaoxin WP1-6 Dong, Enzeng WA1-P

Dong, Enzeng WA1-P Eshaghi, Armin WP2-2 Dong, Enzeng WA1-P -F- Dong, Enzeng TA1-2 Dong, Enzeng TP2-2 Faied, Mariam WP2-1 Dong, Erbao WP2-2 Fan, Chunhui TA1-3 Dong, Erbao WP3-2 Fan, Kairan WP1-3 Dong, Kangkang WP3-2 Fan, Shaowei WP2-2 Dong, Xiangyu WP3-2 Fan, Xiaoshi TA2-4 Dong, Xinyu FA1-2 Fang, Yuansheng TP1-4 Doshi, Ronak WP3-2 Farrukh, Fasih Ud Din FA1-4 Dou, Jianping WP2-3 Feng, Jiasi WA1-P Du, Dongdong TP1-3 Feng, Jingxiang WP2-1 Du, Dongdong TA2-4 Feng, Mengting WA1-P Du, Sheng WP1-6 Feng, Sheng WA1-P Du, Shengzhi WA1-P Feng, Sheng WA1-P Du, Shengzhi TP2-2 Feng, Suxiang FA2-4 Du, Xintian TP2-1 Feng, Xin WP2-5 Du, Yanping TA1-4 Frank, Tobias WP3-3 Du, Yaxin TA2-2 Frank, Tobias TP1-1 Du, Yuhong WA1-P Fu, Chenglong FA1-4 Du, Zeshuai WA1-P Fu, Jiajun FA2-4 Du, Zeshuai WP2-4 Fu, Jingheng WP2-6 Du, Zhangming WP3-4 Fu, Mingyu WP3-6 Duan, Dengping TP2-1 Fu, Qiang WP1-1 Duan, Xiaoshuai FA1-3 Fu, Qiang TP2-4 Duan, Xiaoshuai FA1-5 Fu, Qiang FA2-2 Dyer, Benjamin Mahlon WA1-P Fu, Qiang FA2-2 -E- Fu, Qiang FA2-4 Fu, Xudong TP2-1 Ejima, Yoshimichi TA1-5 Fu, Yongling WP1-2 Ejima, Yoshimichi TA1-5 Fu, Zhuang WP2-5 Ejima, Yoshimichi TA2-5 -G- Elmkaiel, Ghadeer TP2-3 Emaru, Takanori TP1-2 Gadsden, S. Andrew WA1-P Ermacora, Gabriele FA1-3 Gao, Gaofeng TP2-1

Gao, Pengfei WA1-P Guo, Jian WA1-P Gao, Shiqi WP2-4 Guo, Jian WP1-1 Gao, Wenyang TA2-6 Guo, Jian WP1-1 Gao, Xiangyu WP1-5 Guo, Jian WP1-1 Gao, Xueshan WA1-P Guo, Jian TP2-2 Gao, Xueshan WP1-2 Guo, Jian TP1-3 Gao, Xueshan TP1-5 Guo, Jian TP2-3 Gao, Xueshan FA2-3 Guo, Jian TA1-4 Gao, Xuewei WA1-P Guo, Jian TP2-4 Gao, Xuewei TP1-1 Guo, Jian TA1-6 Gao, Ya WA1-P Guo, Jian TA1-6 Gao, Yi WA1-P Guo, Jian TP1-6 Gao, Yi WA1-P Guo, Jian FA2-1 Ge, Chuanbin WP1-1 Guo, Jian FA2-1 Ge, Weimin WA1-P Guo, Jian FA2-2 Ge, Weimin TA2-2 Guo, Jian FA2-2 Ge, Weimin FA2-4 Guo, Jian FA2-3 Geng, Haipeng WA1-P Guo, Jian FA1-4 Geng, Haipeng WA1-P Guo, Jian FA1-4 Geng, Haipeng WP3-5 Guo, Jian FA2-4 Geng, Haipeng FA1-5 Guo, Jian FA2-4 Geng, Haipeng FA1-5 Guo, Jin FA2-2 Geng, Yanjuan FA2-4 Guo, Kangkai TA2-3 Giebert, Dennis TA2-1 Guo, Lanjie WP1-5 Go, Ritsu TA2-5 Guo, Mingquan TP2-6 Gold, Tobias WP2-2 Guo, Rui WP2-2 Goller, Tim WP2-2 Guo, Shuxiang WA1-P Gou, Zhenyu WP1-2 Guo, Shuxiang WA1-P Graichen, Knut WP2-2 Guo, Shuxiang WA1-P Gu, Feng TP2-1 Guo, Shuxiang WP1-1 Gu, Weiqun TA2-4 Guo, Shuxiang WP1-1 Guan, Hongyao FA1-4 Guo, Shuxiang WP1-1 Guan, Ping FA1-5 Guo, Shuxiang WP3-2 Guan, Yisheng WP3-2 Guo, Shuxiang WP3-4 Guan, Yueming TP2-1 Guo, Shuxiang WP1-6

Guo, Shuxiang WP3-6 He, Lixin FA2-1 Guo, Shuxiang WP3-6 He, Xiangkun WP1-2 Guo, Shuxiang TP2-2 He, Yuqing TP2-1 Guo, Shuxiang TA1-3 Hirata, Yasuhisa TA1-1 Guo, Shuxiang TA2-3 Hong, Woo Cheol WP3-5 Guo, Shuxiang TP1-3 Hou, Chenggang FA1-5 Guo, Shuxiang TP1-3 Hou, Jianfeng TP1-4 Guo, Shuxiang TP2-3 Hou, Xihuan WP1-6 Guo, Shuxiang TP2-3 Hou, Xihuan WP3-6 Guo, Shuxiang TA1-4 Hou, Xihuan TA2-3 Guo, Shuxiang TP1-4 Hou, Xihuan TP2-3 Guo, Shuxiang TP2-4 Hou, Xihuan TP1-4 Guo, Shuxiang TP2-4 Hou, Xihuan FA2-1 Guo, Ting TA2-5 Hsu, Fang-Chi WP2-3 Guo, Xiang TP1-1 Hu, Bin TP2-5 Guo, Yangming WP3-2 Hu, Fan TA1-1 Guo, Yingwei WA1-P Hu, Haifeng TP2-5 Guo, Zhicheng FA1-2 Hu, Jiaxiang WA1-P Guo, Zhiling TP2-2 Hu, Jiejun WP1-5 Guo, Zixie WP1-4 Hu, Kun TP2-1 -H- Hu, Ming TP2-6 Hu, Pengfei WA1-P Habib, Maki K. WP2-6 Hu, Ruiqin WP1-3 Han, Shujun TP1-5 Hu, Yao WP1-6 Han, Yali FA1-4 Hu, Yao WP3-6 Han, Zejie TP2-6 Hu, Yaqi FA2-2 Han, Zhu FA1-5 Hu, Zhenxuan FA1-2 Han, Zhu FA1-5 Hua, Xiaodong TP1-2 Hao, Minghui FA1-3 Huang, Fang FA2-2 Hao, Minghui FA1-5 Huang, Jian TA1-5 Hao, Yufei FA1-3 Huang, Jie WP3-2 He, Haocheng WA1-P Huang, Nieyong WP2-3 He, Haocheng TP1-1 Huang, Nieyong WP3-5 He, Jialin WP2-6 Huang, Nieyong WP3-5 He, Linglong WP2-4 Huang, Qiang FA1-2

Huang, Qiang FA1-4 Jiang, Yanbing FA2-4 Huang, Xiaobo TP1-4 Jiang, Yancao FA1-4 Huang, Yu WP3-3 Jiang, Zhicheng WA1-P Huang, Yu FA1-3 Jiang, Zijian WP3-6 Huang, Zeyu WA1-P Jin, Haoqiang WP2-1 Huang, Zeyu WA1-P Jin, Kui FA1-2 Huang, Zeyu TA1-4 Jin, Lingyan TP2-6 Huang, Zeyu FA1-1 Jin, Minghe WP1-5 Huo, Xiaojie WP1-1 Jin, Xiaoliang TP1-6 Huo, Xijian FA1-3 Jin, Xiaoliang TP1-6 Huo, Xijian FA1-5 Jing, Cunlei WA1-P Hwang, Hui Geon WP3-5 Jing, Litao WA1-P -I- -K- Idowu, Oluwagbenga Paul FA2-4 Kaczor, Daniel TA1-1 Imai, Shinichi WP2-4 Kan, Haoxuan TA1-3 Inui, Shota TP2-2 Kanda, Shinsuke WP2-6 -J- Kang, Jianbing WP1-5 Kang, Tong FA1-1 Jee, Myeong Jun WP3-5 Kang, Xin TP2-2 Ji, Xiaodong FA1-1 Kang, Yan WA1-P Ji, Xiaoyu WA1-P Karki, Hamad WP1-2 Ji, Xuewu WP1-2 Kim, Hong Rok WP3-5 Jia, Chang TP2-3 Kira, Ryusei WP2-6 Jia, Chao WA1-P Kobayashi, Yukinori TP1-2 Jia, Chao TP1-1 Kobler, Jan-Philipp WA1-P Jia, Jieyu WP1-2 Kong, Jianyi FA1-3 Jia, Tong TA2-3 Kong, Shihan TP2-3 Jia, Wenchuan FA2-1 Kosaki, Takahiro WP1-4 Jiang, Haowen FA1-3 -L- Jiang, Kuncheng TP2-2 Jiang, Li WP2-2 Lai, Juezhu TP1-5 Jiang, Naifu FA2-4 Lai, Wenxin TP1-5 Jiang, Song FA2-5 Le, Hoai Nam FA2-2 Jiang, Xudong TA2-4 Lee, Chien-Chia WP1-4

Lei, Yayun TP1-2 Li, Jinhua FA1-2 Leng, Jianwei TP2-2 Li, Jinhua FA1-2 Leng, Jianwei FA2-1 Li, Jinxiang TA2-2 Li, Aiping WP2-3 Li, Jinyu WA1-P Li, Bichun TA1-3 Li, Juan TP1-4 Li, Bilin WA1-P Li, Juan TP1-4 Li, Bin WA1-P Li, Juan TP2-4 Li, Bin WA1-P Li, Junyuan TA1-5 Li, Bin WP1-3 Li, Kun TP2-5 Li, Bin FA2-2 Li, Leihui WP1-4 Li, Bingjue TP2-5 Li, Liangjie TP2-4 Li, Bo WA1-P Li, Lixin FA1-1 Li, Bo TA2-4 Li, Peng WP3-1 Li, Chenming TA1-2 Li, Peng WP2-4 Li, Chensheng TP2-4 Li, Pengyu TA1-6 Li, Chongyang WP2-2 Li, Pengyun FA1-2 Li, Chuming TP2-5 Li, Qi TP2-1 Li, Chunlin TA2-5 Li, Qi TA1-5 Li, En WP2-2 Li, Qin WP1-1 Li, Feng WA1-P Li, Ruibing WA1-P Li, Guanglin FA2-4 Li, Shigang WP1-4 Li, Guoyuan TA1-1 Li, Shuai WA1-P Li, Hongyan TA1-2 Li, Shuangyang WP3-3 Li, Hui TP2-4 Li, Wenqiang WP2-1 Li, Ji TP2-2 Li, Wenqiang WP1-4 Li, Ji TP1-3 Li, Wensheng TP1-4 Li, Ji TA2-4 Li, Xiangxin FA2-4 Li, Jian WA1-P Li, Xiaobin FA1-3 Li, Jian WP1-2 Li, Xin TP2-5 Li, Jian TP1-5 Li, Xiuzhi TA2-3 Li, Jian FA2-3 Li, Xue FA2-3 Li, Jianmin FA1-2 Li, Xuesong FA2-5 Li, Jianmin FA1-2 Li, Xun FA2-4 Li, Jing WP1-5 Li, Yanjun TA2-6 Li, Jingyan TA2-4 Li, Yao FA1-3

Li, Yibin FA2-3 Lin, Wen-Jie WP2-3 Li, Yibin FA2-3 Lin, Yue TP1-5 Li, Yinlong WP1-5 Lin, Yuqing FA1-2 Li, Youfu WP1-4 Lin, Zhe WP3-1 Li, Yujia WA1-P Lin, Zhe WP3-1 Li, Yujia WA1-P Lin, Ziwang WP1-2 Li, Yujia TA1-4 Lin, Ziyan TP2-4 Li, Yujia FA1-1 Ling, Yihong TP2-5 Li, Yunpeng FA1-5 Liu, Bin WP1-5 Li, Zan WP1-6 Liu, Bingchen WP2-2 Li, Zan WP3-6 Liu, Bohao WA1-P Li, Zan TA2-3 Liu, Changbo TP2-1 Li, Zan TP2-3 Liu, Chun WP1-3 Li, Zan TP1-4 Liu, Chun FA2-2 Li, Zan FA2-1 Liu, Dekun TA1-1 Li, Zhaotong TP1-5 Liu, Fanming WP3-3 Li, Zhaoyang WP1-5 Liu, Fengyu FA1-2 Li, Zhaoyang FA1-1 Liu, Gangfeng TP1-4 Li, Zhihan WP3-6 Liu, Guangjun TA1-3 Li, Zhihui TP1-5 Liu, Guanjun TP1-4 Li, Zhiqi FA1-1 Liu, Hai WP1-4 Li, Zhiyuan WP3-1 Liu, Hanze TP1-3 Li, Zhongbin TP2-1 Liu, Hao TA1-4 Li, Zhonghua TP2-5 Liu, Hao TP2-6 Liang, Bin WP3-2 Liu, Hong WP1-5 Liang, Huawei WP3-1 Liu, Hongli TP2-2 Liang, Huawei TA1-3 Liu, Hongli TP1-3 Liang, Jingyuan FA1-5 Liu, Hongli TA2-4 Liang, Liyuan FA1-1 Liu, Houde WP3-2 Liang, Xingcan WA1-P Liu, Huan WA1-P Liang, Xingcan WP3-5 Liu, Huan WP1-2 Liang, Xingcan FA2-1 Liu, Huan TP1-5 Lin, Hao WA1-P Liu, Huan FA2-3 Lin, Hao TA1-1 Liu, Jia WA1-P Lin, Hsien-I WP1-4 Liu, Jincun WP3-6

Liu, Jinfu WP3-5 Liu, Yan TP2-1 Liu, Jinfu WP1-6 Liu, Yang TA2-5 Liu, Jinfu FA2-1 Liu, Yanhong WP1-1 Liu, Juan WP1-1 Liu, Yaning WP3-1 Liu, Jun WA1-P Liu, Yanjie TA2-3 Liu, Jun FA1-5 Liu, Yi TA1-6 Liu, Jun FA1-5 Liu, Yi TA1-6 Liu, Lei WP3-5 Liu, Ying TP1-4 Liu, Lei WP1-6 Liu, Yiwei WP1-5 Liu, Liqiang WP3-1 Liu, Yizhuo WP3-1 Liu, Meng WP1-4 Liu, Yu WP1-6 Liu, Na TA1-2 Liu, Yu WP3-6 Liu, Nan TP1-1 Liu, Yu TP1-4 Liu, Qi WP2-1 Liu, Yu FA2-1 Liu, Qiang WP1-3 Liu, Yue WA1-P Liu, Qing TA1-1 Liu, Yue WP3-3 Liu, Qiyuan FA1-5 Liu, Yue TA1-2 Liu, Rongqiang WP1-5 Liu, Yulong WP1-2 Liu, Shuai TP1-2 Liu, Yulong TA1-5 Liu, Shujian TA1-2 Liu, Yulong TA2-5 Liu, Sinan WP1-6 Liu, Yuming WA1-P Liu, Siqi WA1-P Liu, Yunqing WA1-P Liu, Tao TP1-4 Liu, Zhaoyi TP2-4 Liu, Tianlin TA1-1 Lomakin, Alexander WP2-2 Liu, Xiangwei WP1-4 Long, Bo WA1-P Liu, Xiangwei TA1-2 Long, Bo TP1-1 Liu, Xiao TP2-5 Long, Gaoxiang WA1-P Liu, Xiaofo WP2-4 Lu, Hui TP1-2 Liu, Xiaojun WP1-3 Lu, Kuan TP2-2 Liu, Xiaoming FA1-2 Lu, Sixi WP2-2 Liu, Xin WP1-5 Lu, Xiao WP1-1 Liu, Xiufeng WP3-3 Lu, Yan WP2-2 Liu, Xiufeng FA1-3 Luan, Nan WP2-1 Liu, Xuemei WP2-3 Luo, Chen WP2-3 Liu, Yahui WP1-2 Luo, Chen WP2-4

Luo, Chen TA1-2 Ma, Xihan WP2-1 Luo, Dingji WA1-P Ma, Youchun WA1-P Luo, Dingji WP1-2 Ma, Youchun TP1-6 Luo, Dingji TP1-5 Ma, Youchun TP2-6 Luo, Dingji FA2-3 Ma, Youchun FA2-2 Luo, Dingsheng TA1-1 Ma, Youjie TA2-1 Luo, Hanrong WP1-3 Mahadevan, Anita WP3-2 Luo, Man TP1-2 Malek Azari, Mina TA1-1 Luo, Xiangyu FA1-2 Mao, Chenjian WA1-P Lv, Hao WA1-P Mao, Peiye WA1-P Lv, Jiadong WP1-6 Mao, Wujun FA2-1 Lv, Jiadong WP3-6 Mechali, Abdesslam TA2-3 Lv, Jiale WA1-P Meng, Fanqiang TP2-1 Lv, Jiale WP1-2 Meng, Hao TP2-1 Lv, Jiale TP1-5 Meng, Xuehao WA1-P Lv, Jiale FA2-3 Miao, Shuaitong WA1-P Lv, Lianrong TP1-2 Miao, Wei TA2-4 Lv, Pengfei WA1-P Miki, Kohei WP2-6 Lv, Pengfei WP1-2 Mills, James K. WP2-2 Lv, Pengfei TP1-5 Mills, James K. WP3-4 Lv, Pengfei FA2-3 Ming, Aiguo FA1-4 Lyu, Chuqiao TP1-6 Ming, Liangjie TP1-1 Lyu, Chuqiao TP2-6 Ming, Liangjie TP2-5 Lyu, Chuqiao FA2-2 Moriya, Naoki TP2-3 Lyu, Dingchong FA1-5 Mou, Qisong TA2-2 -M- -N- Ma, Baoping TA1-3 Nagai, Isaku WP2-6 Ma, Bing FA1-3 Nagai, Isaku WP2-6 Ma, Hongbin TP1-2 Nagai, Isaku TA1-1 Ma, Jianhua WP3-3 Nagata, Fusaomi WP2-6 Ma, Jianhua FA1-3 Nakata, Sotaro WP2-6 Ma, Liangping WP1-6 Ngo, Hoang Phuc Van FA2-2 Ma, Shugen FA2-1 Ni, Jianyun WA1-P Ma, Tianyi TA1-4 Ni, Jianyun TP1-3

Ni, Shiquan TA1-4 -Q- Ning, Hui TP1-5 Nishide, Shun TP2-2 Qi, Yong WA1-P Niu, Hang FA1-5 Qi, Yong TP1-3 Niu, Jiaxi TA2-6 Qiao, Dan TA1-4 Niu, Jundao TP1-5 Qiao, Haiyan TP2-1 Niu, Shunkang FA1-2 Qiao, Yifan TA1-2 Norris, William R. WP2-1 Qin, Juan TP1-2 -O- Qin, Juan TP2-2 Qin, Mingyue FA1-4 Omar, Mechali TA2-3 Qin, Xiansheng TP1-2 Ortmaier, Tobias WA1-P Qiu, Binbin TP1-1 Ortmaier, Tobias WP3-3 Qiu, Binbin TP2-5 Ortmaier, Tobias WP2-5 Qiu, Binbin TP2-5 Ortmaier, Tobias WP2-5 Qiu, Changlin TP2-3 Ortmaier, Tobias TA1-1 Qiu, Chenguang TP2-1 Ortmaier, Tobias TA2-1 Qiu, Jing TP1-4 Ortmaier, Tobias TP1-1 Qiu, Shi TA1-2 Otsuka, Akimasa WP2-6 -R- -P- Rahimi, Fariba WP3-5 Pan, Jie WP3-6 Ram Sankar, Arvind WP3-2 Pan, Qinxue WP3-3 Rao, Madhav WP3-2 Pan, Qunhua TP2-1 Ravankar, Ankit A. TP1-2 Pan, Ruipeng WP3-3 Reddy, Aravind WP3-2 Pang, Lei WP3-1 Reddy, Sunil Kumar WP1-2 Park, Sang Hyeok WP3-5 Ren, Fuji TP2-2 Pei, Ling FA1-3 Ren, Fuji FA1-1 Peng, Tongtong WP3-5 Ren, Hao TP2-5 Peng, Xiuyan TA2-5 Ren, Keguang WA1-P Perner, Lars WP2-5 Ren, Weijia WA1-P Perner, Lars WP2-5 Ren, Yanna TA1-5 Pham, Anh Duc FA2-2 Ren, Yanna TA1-5 Piao, Yan TA1-3 Ren, Yanna TA2-5 Popp, Eduard TA2-1 Rong, Jikun WP2-1

Rovasenda, Manfredi FA1-3 Shi, Jia WA1-P -S- Shi, Jia WP3-5 Shi, Jia WP1-6 Salazar Luces, Jose Victorio TA1-1 Shi, Jinfei FA1-4 Samuel, Oluwarotimi Williams FA2-4 Shi, Liwei WP1-6 Sartori, Daniele FA1-3 Shi, Liwei WP3-6 Sawada, Hideyuki TP2-3 Shi, Liwei TA2-3 Sawada, Hideyuki FA2-2 Shi, Liwei TP2-3 Sawada, Hideyuki FA2-4 Shi, Liwei TP1-4 Schappler, Moritz TA1-1 Shi, Liwei FA2-1 Senouci, Abdelkader TA2-3 Shi, Peng TP1-5 Serebrenny, Vladimir Valeryvich TP2-3 Shi, Peng TP1-6 Shang, Lamei TP2-1 Shi, Peng TP1-6 Shang, Peng WP2-3 Shi, Peng FA1-4 Shang, Peng WP3-5 Shi, Yangyang WP3-3 Shang, Peng WP3-5 Shi, Yankai TP2-5 Shang, Weiwei WP3-2 Shi, Yongjie WA1-P Shao, Haicun TA1-3 Shi, Yongjie WP1-2 Shao, Lei TP2-2 Shi, Yongjie TP1-5 Shao, Lei TP1-3 Shi, Yongjie FA2-3 Shao, Lei TA2-4 Shi, Yuhua WP2-3 Shao, Xin WA1-P Shi, Yuhua WP3-5 Shao, Xingmao TP1-1 Shi, Yuhua WP3-5 Shen, Feng WP2-1 Shi, Yunde WP1-6 Shen, Feng WP1-4 Shi, Yunde WP3-6 Shen, Jian TP1-5 Shi, Yunling TA2-5 Shen, Linjie WP1-2 Shigemune, Hiroki TP2-3 Shen, Rui TP2-6 Shigemune, Hiroki FA2-2 Shen, Yichao WP2-2 Shuxiang, Guo TA1-6 Shen, Zhenghui FA2-5 Shuxiang, Guo TA1-6 Shi, Chaoyang FA1-2 Shuxiang, Guo TA1-6 Shi, Chaoyang FA1-2 Shuxiang, Guo TA1-6 Shi, Chaoyang FA1-2 Shuxiang, Guo TA2-6 Shi, Chaoyang FA1-2 Shuxiang, Guo TP1-6 Shi, Chuanqi FA1-4 Shuxiang, Guo TP1-6

Shuxiang, Guo TP1-6 Su, Liying TP1-3 Shuxiang, Guo TP1-6 Sui, Wenbo WP3-6 Shuxiang, Guo TP2-6 Sui, Wenbo TA1-3 Shuxiang, Guo TP2-6 Sui, Wenbo TP1-3 Shuxiang, Guo TP2-6 Sun, Han FA1-4 Shuxiang, Guo FA2-1 Sun, Hanxu WP3-2 Shuxiang, Guo FA2-1 Sun, Honglin WP2-1 Shuxiang, Guo FA2-1 Sun, Jie TP2-5 Shuxiang, Guo FA2-2 Sun, Kui WP1-5 Shuxiang, Guo FA2-2 Sun, Lin TA2-4 Shuxiang, Guo FA2-2 Sun, Liting WP1-2 Shuxiang, Guo FA2-3 Sun, Liwei TA2-5 Shuxiang, Guo FA1-4 Sun, Pengpeng WP1-3 Shuxiang, Guo FA1-4 Sun, Qi WP2-3 Shuxiang, Guo FA2-4 Sun, Wu FA2-1 Shuxiang, Guo FA2-4 Sun, Xiangyu TP1-4 Shu, Ruizhi WP2-5 Sun, Xiuyu TP2-4 Shuai, Liguo WA1-P Sun, Yanan FA1-2 Shuai, Liguo TA1-3 Sun, Yongjun WP1-5 Shukla, Amit WP1-2 Sun, Yue TA1-4 Si, Guoning WP3-4 Sun, Yutong TP1-5 Singh, Christopher TA1-3 Sun, Zhenyu TA2-4 Skulstad, Robert TA1-1 -T- Smith, Trevor Robin WA1-P Song, Dapeng TP1-6 Takahashi, Satoshi TA1-5 Song, Dapeng TP1-6 Takahashi, Satoshi TA1-5 Song, Jiaorong WP2-3 Takahashi, Satoshi TA2-5 Song, Li WP1-5 Tan, Fulong WP3-4 Song, Sifan WP3-6 Tan, Rulong WP2-5 Song, Wei WP2-5 Tan, Yingzi WP2-1 Song, Ze WP3-4 Tan, Yingzi TP2-4 Su, Chun WP1-3 Tang, Di WA1-P Su, He FA1-2 Tang, Di WA1-P Su, Jionglong WP3-6 Tang, Di TA1-4 Su, Liying TA2-2 Tang, Di FA1-1

Tang, Peng WP1-5 -W- Tang, Wenxian WP1-6 Tang, Xiaoqing FA1-2 Wang, Changsong TP2-5 Tang, Xiaoya TP1-3 Wang, Chenglong FA1-2 Tang, Xiaoying TA2-6 Wang, Chun WP1-5 Tang, Xiaoying TA2-6 Wang, Chunjie WA1-P Tang, Xuming WP2-2 Wang, Chunlei TA1-3 Tang, Zixin WP1-5 Wang, Cong WP1-3 Tantau, Mathias WP2-5 Wang, Cong FA2-2 Tantau, Mathias WP2-5 Wang, Gang TA2-4 Tantau, Mathias TA1-1 Wang, Guishan TP1-4 Tantau, Mathias TA2-1 Wang, Hainan WA1-P Tao, Chunjing FA1-4 Wang, Hainan TP1-3 Tao, Meng WP3-2 Wang, Hao WP2-1 Tendeng, Awa WP3-6 Wang, Hongjun WA1-P Thanh, Vo Nhu FA2-4 Wang, Hongjun WP3-3 Tian, Baoqiang WP1-6 Wang, Hongjun TA2-2 Tian, Dapeng TA1-4 Wang, Hongjun TA2-2 Tian, Jiale WA1-P Wang, Hongjun FA1-1 Tian, Jing WP3-4 Wang, Hongjun FA1-1 Tian, Xiaoli TP1-5 Wang, Hongjun FA1-1 Tian, Yujia TP1-5 Wang, Hongjun FA2-5 Tomizuka, Masayoshi WP1-2 Wang, Hongjun FA2-5 Tong, Ru WP1-2 Wang, Huaixin WP1-1 Tse, Peter U. TA2-5 Wang, Huan WA1-P -U- Wang, Huixin TP2-4 Wang, Jian WP1-3 Uhlig, Frank TP2-5 Wang, Jianjian FA1-3 -V- Wang, Jing TP1-5 Wang, Junyao FA1-1 Vazhiyal, Vikas WP3-2 Wang, Kaidi TP1-6 Vo, Nhu Thanh FA2-2 Wang, Kaidi TP1-6 Vo, Quoc Van FA2-2 Wang, Ke TP1-4 Volkmann, Björn TA1-1 Wang, Kejin WP2-5 Völz, Andreas WP2-2 Wang, Keqing TA1-3

Wang, Lei WA1-P Wang, Tao TA1-5 Wang, Lei TA1-4 Wang, Taowei TP1-4 Wang, Lei FA2-5 Wang, Tianzhu WP1-2 Wang, Liutao WA1-P Wang, Tongtong TA1-1 Wang, Manyi TA1-2 Wang, Wei TP1-4 Wang, Min TA2-4 Wang, Weihua TA1-5 Wang, Mingkang WP1-2 Wang, Wenbo FA1-3 Wang, Qi WA1-P Wang, Xiaofei TP1-5 Wang, Qianjin WP1-6 Wang, Xiaofeng WA1-P Wang, Qin TP1-4 Wang, Xiaofeng TA2-2 Wang, Rentao FA2-3 Wang, Xin TP1-2 Wang, Riwei WP1-4 Wang, Xingdong FA1-3 Wang, Sen WA1-P Wang, Xinming WA1-P Wang, Sen WA1-P Wang, Xinming TA1-4 Wang, Sen WA1-P Wang, Xinrui WP1-1 Wang, Sen TA1-4 Wang, Xue WP2-4 Wang, Sen FA1-1 Wang, Xuemei FA2-4 Wang, Shaobo WP3-1 Wang, Xuemei FA2-5 Wang, Shaobo TA1-3 Wang, Yahao WP2-2 Wang, Shibei WP2-3 Wang, Yahao WP3-2 Wang, Shibei WP3-5 Wang, Yalong TP1-1 Wang, Shigang WA1-P Wang, Yalong TP1-5 Wang, Shigang WP1-2 Wang, Yanen WA1-P Wang, Shigang WP2-5 Wang, Yangyang WA1-P Wang, Shigang TP1-5 Wang, Yao WP2-5 Wang, Shigang FA2-3 Wang, Yaohong FA1-3 Wang, Shoujun TP1-1 Wang, Yining FA1-3 Wang, Shuo WP3-4 Wang, Yu TP2-5 Wang, Shuo TA1-2 Wang, Yue TP1-6 Wang, Shuoyu FA1-4 Wang, Yue TP2-6 Wang, Shuxin FA1-2 Wang, Yue FA2-2 Wang, Shuxin FA1-2 Wang, Yuebo WA1-P Wang, Sining TP2-2 Wang, Yuezong WP3-4 Wang, Su TP1-2 Wang, Yujing WP2-5 Wang, Tao TA1-5 Wang, Yunhe TA1-4

Wang, Yunliang WA1-P Wu, Haiyuan WP2-4 Wang, Yunliang WA1-P Wu, Han WP1-2 Wang, Yunliang WA1-P Wu, Hanhui WA1-P Wang, Zhen WP1-3 Wu, Hong WA1-P Wang, Zhen FA2-2 Wu, Hong WP1-3 Wang, Zhihai WP3-4 Wu, Hong TP1-2 Wang, Zhihai FA1-3 Wu, Jin WA1-P Wang, Zhisen TA2-6 Wu, Jinglong TA1-5 Wang, Zhiyong TP1-2 Wu, Jinglong TA1-5 Wang, Ziqing WP3-5 Wu, Jinglong TA2-5 Wang, Zixu WP3-4 Wu, Jinglong TA2-5 Wang, Zixu TP2-4 Wu, Qiong TA1-5 Watanabe, Keigo WP2-6 Wu, Qiong TA1-5 Watanabe, Keigo WP2-6 Wu, Qiong TA2-5 Watanabe, Keigo WP2-6 Wu, Qiong TA2-5 Watanabe, Keigo TA1-1 Wu, Ruolin WP3-3 Wei, Hongmiao TA2-2 Wu, Xiangfei WP2-4 Wei, Jiaqi WA1-P Wu, Xiangfei FA2-3 Wei, Jiaqi TP1-1 Wu, Xiaoyong WP2-5 Wei, Mingzhu FA2-3 Wu, Yanjuan WA1-P Wei, Wei TP2-4 Wu, Yaowu TA1-4 Wei, Yanhui WP2-6 Wu, Zhengxing WP1-6 Wei, Yaoyao TA1-1 Wu, Ziyu TP2-5 Wei, Yonggeng WA1-P -X- Wei, Zikang WA1-P Wen, Xiaodong WA1-P Xi, Minglei WA1-P Wen, Yubin TP2-1 Xi, Pengyu FA2-4 Wielitzka, Mark WA1-P Xia, Binyuan TA2-1 Wielitzka, Mark WP3-3 Xia, Dan WP1-6 Wielitzka, Mark WP2-5 Xia, Dan WP3-6 Wielitzka, Mark WP2-5 Xia, Debin WP1-6 Wielitzka, Mark TA2-1 Xia, Debin WP3-6 Wielitzka, Mark TP1-1 Xia, Debin TA2-3 Wu, Dongrui TA1-5 Xia, Debin TP2-3 Wu, Duidi TP1-1 Xia, Debin TP1-4

Xia, Debin FA2-1 Xu, Jigang WA1-P Xia, Guoqing TA2-1 Xu, Jigang WP1-1 Xia, Zhengbing FA1-3 Xu, Jigang TP1-3 Xiang, Zhaowei WP2-5 Xu, Jigang TP2-3 Xiao, Nan TP1-6 Xu, Jigang FA2-3 Xiao, Nan TP1-6 Xu, Liang TP1-1 Xiao, Tao WP1-5 Xu, Liang TP1-5 Xie, Bang WA1-P Xu, Limei TA2-3 Xie, Jia WP3-2 Xu, Lin TP1-5 Xie, Xiang FA1-4 Xu, Linsen WA1-P Xie, Xiaomei WP2-4 Xu, Linsen WP3-5 Xie, Xiaomei FA2-3 Xu, Linsen WP1-6 Xie, Xiomei TA2-3 Xu, Linsen FA2-1 Xin, Yi WP1-1 Xu, Ping WP2-6 Xin, Yi WP1-1 Xu, Ping WP2-6 Xing, Huiming WP1-6 Xu, Ruihan FA1-3 Xing, Huiming WP3-6 Xu, Shuo WA1-P Xing, Huiming TA2-3 Xu, Xiaoming FA2-4 Xing, Huiming TP2-3 Xu, Xiaoming FA2-5 Xing, Huiming TP1-4 Xu, Xiaoyu WP3-3 Xing, Huiming FA2-1 Xu, Xinyue FA1-2 Xing, Yuan FA1-2 Xu, Xiongshi TA1-1 Xing, Yunlong TP2-3 Xu, Yingqiu WP2-1 Xiong, Huilin TP1-2 Xu, Yingqiu TP2-4 Xiong, Mingkang TP1-2 Xu, Yujie WP3-6 Xu, Chao WP2-3 Xu, Zhi TA2-6 Xu, Chunquan WP2-6 Xue, Yongjiang TP1-5 Xu, Chunquan WP2-6 -Y- Xu, Enwei WP2-1 Xu, Hong WP3-5 Yan, Hongkui WA1-P Xu, Hong WP1-6 Yan, Hongkui WA1-P Xu, Hong FA2-1 Yan, Hongkui FA1-1 Xu, Jiajun WA1-P Yan, Nanxing WP3-1 Xu, Jiajun WP3-5 Yan, Ran WP2-5 Xu, Jiajun WP1-6 Yan, Xiang WP3-3

Yan, Xiaolong TP1-5 Yang, Xianglin TA2-6 Yan, Xuejiao FA2-5 Yang, Ya TA1-3 Yang, Bai FA1-1 Yang, Yingjian WA1-P Yang, Baisong WA1-P Yang, Yuan TA1-2 Yang, Baisong WA1-P Yang, Ziyi TA1-6 Yang, Baisong TA1-1 Yang, Ziyi TA1-6 Yang, Baojun WP3-4 Yao, Jiayi WA1-P Yang, Cheng WA1-P Yao, Jiayi FA1-5 Yang, Cheng WP3-2 Yao, Yao WP2-1 Yang, Cheng TP2-6 Yao, Yihui WP2-2 Yang, Chengyu TP1-4 Ye, Xiufen WA1-P Yang, Guang FA1-4 Yi, Xueyong FA1-3 Yang, Hongbo TA2-6 Yi, Yang WP1-3 Yang, Jiajia TA1-5 Yin, Boqun WP2-1 Yang, Jiajia TA1-5 Yin, He WP1-6 Yang, Jiajia TA2-5 Yin, He WP3-6 Yang, Jing TP2-6 Yin, He TA2-3 Yang, Junyou FA1-4 Yin, Jinliang WA1-P Yang, Lei WP1-1 Yin, Tuo FA1-3 Yang, Liusong TP1-1 Yin, Xiliang WP3-5 Yang, Mengke WP2-3 Yin, Zhenshuo WP1-3 Yang, Mengke WP3-3 Yokota, Takayoshi FA2-3 Yang, Miao FA2-3 Yoshida, Kazushi WP2-6 Yang, Min TP1-1 Yu, Biao TA1-3 Yang, Min TP2-5 Yu, Chenglong TP2-5 Yang, Min TP2-5 Yu, Hong TA1-4 Yang, Min TP2-5 Yu, Hongtao TA1-5 Yang, Ruijin WP3-2 Yu, Jia TA1-5 Yang, Run WP2-2 Yu, Junzhi WP3-1 Yang, Run WP3-2 Yu, Junzhi WP1-2 Yang, Shuai TA1-6 Yu, Junzhi WP1-6 Yang, Wei WP3-2 Yu, Junzhi WP3-6 Yang, Weiping TA1-5 Yu, Junzhi TP2-3 Yang, Weiping TA1-5 Yu, Lie WA1-P Yang, Wenhao WP3-3 Yu, Lie WA1-P

Yu, Lie TA1-1 Zeng, Qingshan WP1-1 Yu, Qingguang WA1-P Zeng, Tonghui WP2-4 Yu, Tao TA1-1 Zeng, Tonghui FA2-3 Yu, Tao TP2-6 Zeng, Yuwen TP1-6 Yu, Wenxian FA1-3 Zeng, Yuwen TP1-6 Yu, Xiaohang WP3-1 Zeng, Zhaoheng WP3-2 Yu, Yinghua TA2-5 Zhai, Xiang TP1-1 Yu, Yiyang TA1-5 Zhai, Xiang TP1-5 Yu, Yiyang TA2-5 Zhan, Minghao TA1-3 Yu, Zhangguo FA1-4 Zhang, Ai WP3-1 Yu, Zongxin FA1-3 Zhang, Ai WP1-5 Yuan, Bixin TA1-5 Zhang, Anyuan TA1-5 Yuan, Hang TP1-6 Zhang, Baofeng TA1-2 Yuan, Hang TP1-6 Zhang, Baofeng TP2-3 Yuan, Jianjun FA2-1 Zhang, Bin WA1-P Yuan, Wei TP2-5 Zhang, Canran WP2-3 Yuan, Wenpeng TA1-3 Zhang, Chong FA1-2 Yuan, Xiaodong WA1-P Zhang, Chun FA1-4 Yue, Chao WA1-P Zhang, Fei WP3-2 Yue, Youjun WA1-P Zhang, Fujun WA1-P Yue, Youjun WP3-3 Zhang, Gang WP2-3 Yue, Youjun TA2-2 Zhang, Gang WP2-4 Yue, Youjun TA2-2 Zhang, Gang TA1-2 Yue, Youjun FA1-1 Zhang, Guokai FA1-2 Yue, Youjun FA1-1 Zhang, Guokai FA1-2 Yue, Youjun FA1-1 Zhang, Guokai FA1-2 Yue, Youjun FA2-5 Zhang, Hao WP2-2 Yue, Youjun FA2-5 Zhang, Houxiang TA1-1 Yue, Zhenjia WP1-6 Zhang, Hua TP2-6 Yun, Seongyeol TA2-5 Zhang, Huipeng TA2-2 -Z- Zhang, Jiafan WP2-5 Zhang, Jiahua WP2-3 Zang, Chenxin WP1-3 Zhang, Jiaming WP2-1 Zeipel, Henrik TP1-1 Zhang, Jiaming WP3-6 Zeng, Kun FA1-3 Zhang, Jian WA1-P

Zhang, Jianhua TP2-1 Zhang, Shuo TP2-4 Zhang, Jinyu FA2-4 Zhang, Taifeng TP2-5 Zhang, Jinyu FA2-5 Zhang, Tao WP3-2 Zhang, Jiwei WP2-4 Zhang, Tao TA1-1 Zhang, Junxia TA1-6 Zhang, Tianlu WP3-4 Zhang, Juzhong TA2-6 Zhang, Weimin WP3-1 Zhang, Kaitian TP1-3 Zhang, Weiyi FA1-4 Zhang, Lanyong TA1-3 Zhang, Wenming WP1-5 Zhang, Lanyong TP2-6 Zhang, Wenzhe TA1-3 Zhang, Lijian WP1-3 Zhang, Xi WP1-4 Zhang, Lingyu WA1-P Zhang, Xi TA1-2 Zhang, Lining WA1-P Zhang, Xiangqi TA1-1 Zhang, Liuqing TP2-6 Zhang, Xiangyin TA2-3 Zhang, Longjia WP1-4 Zhang, Xin WP3-4 Zhang, Lu WP3-4 Zhang, Xinhua FA1-3 Zhang, Mingxi TP2-1 Zhang, Xinhua FA1-5 Zhang, Peng TA2-2 Zhang, Xu TA2-5 Zhang, Peng TP1-2 Zhang, Xuping WP1-4 Zhang, Peng TP1-3 Zhang, Xuping WP3-4 Zhang, Peng TA1-5 Zhang, Xuping TA2-2 Zhang, Peng TA1-6 Zhang, Yan WP2-5 Zhang, Pengfei WA1-P Zhang, Yang TP2-6 Zhang, Pengfei WP1-2 Zhang, Yanru WP3-1 Zhang, Pengfei TP1-5 Zhang, Yi WA1-P Zhang, Pengfei FA2-3 Zhang, Ying WA1-P Zhang, Ping TA1-1 Zhang, Ying TA1-5 Zhang, Ping TP2-1 Zhang, Yong TP1-4 Zhang, Pingwei TP2-6 Zhang, Yonggang WP2-1 Zhang, Qiang WP1-1 Zhang, Yonggang WP1-4 Zhang, Qiao TA2-4 Zhang, Yu TP1-2 Zhang, Renchao WP1-3 Zhang, Yu TA2-6 Zhang, Runfeng WP1-6 Zhang, Yu FA1-5 Zhang, Sen FA2-4 Zhang, Yu FA1-5 Zhang, Shengpu WA1-P Zhang, Yuan WA1-P Zhang, Shuo WA1-P Zhang, Yuan TA1-4

83 Zhang, Yue TP2-2 Zhao, Pan WP3-1 Zhang, Yugang WP1-2 Zhao, Peng WP1-2 Zhang, Yuhan TP2-2 Zhao, Peng FA2-3 Zhang, Yunmiao WP3-3 Zhao, Shunli TA2-1 Zhang, Yunong TP1-1 Zhao, Xin TP1-5 Zhang, Yunong TP2-5 Zhao, Xinhua WA1-P Zhang, Yunong TP2-5 Zhao, Xinhua WP2-4 Zhang, Yunong TP2-5 Zhao, Yanfeng WP2-6 Zhang, Zhenhua TP2-1 Zhao, Zhijun WP1-5 Zhang, Zhili FA2-4 Zhao, Zhili FA1-1 Zhang, Zhuo WP3-4 Zhao, Zihao WP1-4 Zhang, Zhuo TA2-2 Zhen, Zhuoheng TP2-5 Zhang, Zufeng WA1-P Zheng, Aiqun WP2-6 Zhang, Zufeng TA1-2 Zheng, Aiqun WP2-6 Zhang, Zufeng FA1-3 Zheng, Guishui TA1-1 Zhao, Da WP3-6 Zheng, Jiangtao TA1-4 Zhao, Di FA1-2 Zheng, Liang WP3-6 Zhao, Haibo TP2-5 Zheng, Liang TA1-3 Zhao, Han TP2-2 Zheng, Liang TP1-3 Zhao, Hui WA1-P Zheng, Lin TA1-4 Zhao, Hui WP3-3 Zheng, Lingling WP3-4 Zhao, Hui TA2-2 Zheng, Wei WA1-P Zhao, Hui TA2-2 Zheng, Wei FA1-5 Zhao, Hui FA1-1 Zheng, Xiande TP1-4 Zhao, Hui FA1-1 Zheng, Xiaokun TA1-3 Zhao, Hui FA1-1 Zheng, Xiaoyan TP1-1 Zhao, Hui FA2-5 Zheng, Zhi WP2-6 Zhao, Hui FA2-5 Zhong, Liqiang TP1-4 Zhao, Jianchang FA1-2 Zhong, Weilin TP1-2 Zhao, Jianchang FA1-2 Zhou, Chao WP3-4 Zhao, Jianghai WP2-4 Zhou, Chao WP1-6 Zhao, Jie TP1-4 Zhou, Chao TP2-3 Zhao, Jinqiang TP2-6 Zhou, Haibo TA1-2 Zhao, Mingyang WA1-P Zhou, Haiyang TA1-4 Zhao, Mingyang TP1-3 Zhou, Junjie WA1-P

Zhou, Lelai FA2-3 Zhu, Qingyu TP1-5 Zhou, Lelai FA2-3 Zhu, Qinling WA1-P Zhou, Mugen WP1-6 Zhu, Taiyun WP3-2 Zhou, Mugen WP3-6 Zhu, Xiaojun WP3-2 Zhou, Mugen TA2-3 Zhu, Ximeng TP1-4 Zhou, Mugen TP2-3 Zhu, Yangyang WP1-6 Zhou, Mugen TP1-4 Zhu, Yangyang WP3-6 Zhou, Mugen FA2-1 Zhu, Yaqiao WA1-P Zhou, Ping FA1-3 Zhu, Yufei WP3-3 Zhou, Shengmin TP2-5 Zhu, Yuqin WA1-P Zhou, Wei TP1-6 Zhu, Yuqin TA1-4 Zhou, Wei TP2-6 Zhuang, Chungang WP2-2 Zhou, Weijie FA1-4 Zhuang, Zixiang FA2-1 Zhou, Wenming TA1-4 Ziaukas, Zygimantas WA1-P Zhou, Xiaoguang WP2-3 Zielinska, Teresa FA2-4 Zhou, Xiaoguang WP3-3 Zou, Guangming FA1-3 Zhou, Xuesong TA2-1 Zhou, Xufeng WP1-6 Zhou, Xufeng WP3-6 Zhou, Yijun WP2-3 Zhou, Yijun WP2-4 Zhou, Yijun TA1-2 Zhou, Yuanyuan TP2-6 Zhou, Zhengxue WP1-4 Zhou, Zhenhua TA1-5 Zhou, Zhuo WP3-5 Zhu, Anshi TA2-4 Zhu, Deiming WP1-2 Zhu, Guodong TP1-4 Zhu, Haifei WP3-2 Zhu, Hui WP3-1 Zhu, Jiangang WA1-P Zhu, Junchao TA1-2 Zhu, Junchao TP2-3 Zhu, Lijing TA1-2

85 Memo

Memo

Council Honorary Chair: Yoshiyuki Kakehi, Kagawa University, Japan Advisory Council Honorary Chairs: IEEE ICMA 2021 T. J. Tarn, Washington University, USA Toshio Fukuda, Beijing Institute of Technology, China Advisory Council Chairs: 2021 IEEE International Conference on Yoshihiro Suenaga, Kagawa University, Japan Tianyou Chai, Northeastern University, China Mechatronics and Automation Kazuhiro Kosuge, Tohoku University, Japan August 8-11, 2021, Takamatsu, Japan A.A. Goldenberg, University of Toronto, Canada Paolo Dario, Scuola Superiore Sant'Anna, Italy Masayoshi Tomizuka, UC Berkeley, USA Mario A. Rotea, University of Massachusetts, USA Ju-Jang Lee, KAIST, Korea Ren C. Luo, National Taiwan University, Taiwan

Yu Yao, Harbin Engineering University, China Yanrong Li, UESTC, China Huadong Yu, CUST, China Simin Li, Guangxi Univ. of Science and Tech. , China Co-sponsors: IEEE Robotics and Automation Society, Kagawa University, Beijing Institute of Technology Qingxin Yang, Tianjin University of Tech., China General Chair: Technical Co-sponsors: RSJ, JSME, SICE, JSPE, TJUT, HIT, HEU, UEC, UESTC, CUST, GUST Shuxiang Guo, Kagawa University General Co-Chairs: William R. Hamel, University of Tennessee, USA Tatsuo Arai , Osaka University, Japan Call for Papers Index Darwin G. Caldwell, Italian Institute of Tech., Italy Tim Lueth, TUM, Germany The 2021 IEEE International Conference on Mechatronics and Automation (ICMA James K. Mills, Univ. of Toronto, Canada 2021) will take place in Takamatsu, Kagawa, Japan from August 8 to August 11, Max Q. -H. Meng, Chinese University of Hong Kong 2021. Takamatsu is a small city located at Sikoku which is the smallest island in 4 Makoto Kaneko, Osaka University, Japan main islands of Japan. Shikoku contains a lot of temples including Zentsu-ji, where Jie Zhao, Harbin Institute of Technology, China Baofeng Zhang, Tianjin Univ. of Technology, China one of the most famous Buddhists, Kukai, was born. Tao Jiang, Chengdu Univ. of Information Tech., China As the host city of ICMA 2021, Takamatsu not only provides the attendees with a Yuxin Zhao, Harbin Engineering University, China great venue for this event, but also an unparalled experience in the Japanese history Masao Arakawa, Kagawa University, Japan through several historical architectures. You are cordially invited to join us at IEEE Program Chairs: ICMA 2021 in Takamatsu. The objective of ICMA 2021 is to provide a forum for Keisuke Morishima, Osaka Univ., Japan Qinxue Pan, Beijing Institute of Technology, China researchers, educators, engineers, and government officials involved in the general Program Co-Chairs: areas of mechatronics, robotics, automation and sensors to disseminate their latest Shingo Kagami, Tohoku University, Japan research results and exchange views on the future research directions of these fields. Hong Zhang, University of Alberta, Canada Hidekuni Takao, Kagawa University, Japan Yili Fu, Harbin Institute of Technology, China The topics of interest include, but not limited to the following: Atsushi Yamashita, University of Tokyo, Japan - Intelligent mechatronics, robotics, biomimetics, automation, control systems, Stefan Byttner, Halmstad University, Sweden - Opto-electronic elements and Materials, laser technology and laser processing Huosheng Hu, University of Essex, U.K - Elements, structures, mechanisms, and applications of micro and nano systems Yuanqing Xia, Beijing Institute of Technology, China - Teleoperation, telerobotics, haptics, and teleoperated semi-autonomous systems Hongbin Zha, Peking University, China Tianmiao Wang, Beihang University, China -Sensor design, multi-sensor data fusion algorithms and wireless sensor networks Hesheng Wang, Shanghai Jiaotong University, China - Biomedical and rehabilitation engineering, prosthetics and artificial organs Jingtai Liu, Nankai University, China - Control system modeling and simulation techniques and methodologies Organizing Committee Chairs: - AI, intelligent control, neuro-control, fuzzy control and their applications Hideyuki Hirata, Kagawa University, Japan Aiguo Ming, U. of Electro-Communications, Japan - Industrial automation, process control, manufacturing process and automation Tutorials/Workshop Chairs: Guangjun Liu, Ryerson University, Canada Contributed Papers: All papers must be submitted in PDF format prepared strictly Keisuke Suzuki, Kagawa University, Japan following the IEEE PDF Requirements for Creating PDF Documents for IEEE Baofeng Gao, Beijing Institute of Technology, China Xplore. The standard number of pages is 6 and the maximum page limit is 8 pages Qiquan Quan, Harbin Institute of Technology, China Invited/Organized Session Chairs: with extra payment for the two extra pages. See detailed instructions in the Kotaro Kimura, Osaka University, Japan conference web site. Kazuhito Yokoi, AIST, Japan Paul Wen, Univ. of Southern Queensland, Australia Organized Sessions: Proposals with the title, the organizers, and a brief statement Hideyuki Sawada, Sawada University, Japan of purpose of the session must be submitted to an OS Chair by April 10, 2021. Dapeng Tian, CIOMP, CAS, China Yang Shi, University of Victoria, Canada Tutorials & Workshops: Proposals for tutorials and workshops that address Shuoyu Wang, Kouchi Institute of Technology, Japan related topics must be submitted to one of the Tutorial/Workshop Chairs by May 1, Junzhi Yu, Institute of Automation, CAS, China 2021. Awards Committee Co-chairs James K. Mills, Univ. of Toronto, Canada Xinkai Chen, Shibaura Institute of Technology, Japan Important Dates: Liwei Shi, Beijing Institute of Technology, China April 10, 2021 Full papers and organized session proposals Publications Chair: May 1, 2021 Proposals for tutorials and workshops Jin Guo, Beijing Institute of Technology, China May 15, 2021 Notification of paper and session acceptance Publicity Chairs: June 1, 2021 Submission of final papers in IEEE PDF format Jinjun Shan, York University, Canada Yangmin Li, Hong Kong Polytechnic Univ., China Xiufen Ye, Harbin Engineering University, China For detailed up-to-date information, please visit the IEEE ICMA conference Daisuke Sasaki, Kagawa University, Japan website at: Finance Chair: Jian Guo, Tianjin University of Engineering, China Local Arrangement Chairs: http://2021.ieee-icma.org Hidenori Ishihara, Kagawa University, Japan Koh Inoue, Kagawa University, Japan Secretariats: X. Jin, R. An, P. Shi, Z. Yang, Kagawa U., Japan

Graduate School of Information Sciences, Tohoku University 6-6-01 Aramaki Aza Aoba, Aoba-ku, Sendai 980-8579, Japan

Scope of the Journal

IJMA is a fully refereed international journal that presents the state-of-the-art research in the area of mechatronics and industrial automation. The intention of IJMA is to provide an international forum to report latest developments from interdisciplinary theoretical studies, computational algorithm development and practical applications. It particularly welcomes those emerging methodologies and techniques which bridge theoretical studies and applications and have significant potential for real-world applications.

Contents IJMA publishes original papers, technical reports, case studies, review papers and tutorials. Special Issues devoted to important topics in advanced mechatronic systems, robotics, control engineering and industrial automation will be published from time to time.

Subject coverage: Suitable topics include, but are not limited to: • Intelligent mechatronics, robotics and biomimetics • Novel and unconventional mechatronic International Journal of systems and robots • Modelling and control of mechatronics Mechatronics systems and robots and Automation • Elements, structures, mechanisms of micro and nano systems • Sensors, wireless sensor networks and multi-sensor data fusion • Biomedical and rehabilitation engineering, prosthetics and artificial organs • AI, neural networks and fuzzy logic in mechatronics and robotics • Industrial automation, process control and networked control systems • Telerobotics, human computer interaction, human-robot interaction

www.inderscience.com/ijma

visit www.inderscience.com Members of the Editorial Board

Editor-in-Chief Prof. Jeffrey Johnson Prof. Shuxiang Guo Open University Kagawa University, UK Japan Prof. Jong-Hwan Kim Executive Editor KAIST Prof. Huosheng Hu Korea, Republic of University of Essex, UK Prof. Chang Liu Northwestern University Editorial Board USA Prof. Martin Adams Nanyang Technological University Prof. Max Meng Singapore Chinese University of Hong Kong Hong Kong Prof. Gursel Alici University of Wollongong Prof. James K. Mills Australia University of Toronto Canada Prof. Hajime Asama University of Tokyo Prof. Aiguo Ming Japan University of Electro-Communications Japan Prof. Darwin G Caldwell Italian Institute of Technology Prof. Nazim Mir-Nasiri Italy Swinburne University of Technology Malaysia Prof. Fabio Celani Sapienza University of Rome Prof. T.J. Tarn Italy Washington University in St. Louis USA Prof. Alexander Ferworn Ryerson University Prof. Hong Zhang Canada University of Alberta Canada Prof. Toshio Fukuda Nagoya University Prof. Jianwei Zhang Japan University of Hamburg Germany Prof. William R. Hamel The University of Tennessee, USA

Prof. Dagui Huang University of Electronic Science and Technology of China, China

visit www.inderscience.com Methods of payment

What the editors say To be completed by all subscribers about IJMA Terms are payment with order. Payment by The new IJMA will cover the area of mechatronics and cheque, banker’s draft or credit card is automation that is currently very important and hot topics. I see that the technology of mechatronics and acceptable. automation will play a major role in developing new technological products. Hence, I am expecting that the Name of subscriber new journal will be a great hit with every researcher who is working on future innovations. It gives a great pleasure to be a part of the IJMA team...... Toshio Fukuda, Nagoya University, Japan ...... I am honored to be a part of this exciting new journal Position ...... which addresses an important and emerging area of research, and believe that it will be extraordinarily successful. I am confident that IJMA will have a Institution ...... significant impact in the area of mechatroronics and automation. Address ...... T.J. Tarn, Washington University in St. Louis, USA ......

It is certainly an honor for me to serve as an Editorial ...... Board member for this important new journal. Mechatronics and automation technologies are central in a wide range of expanding industries across the world...... IJMA promises to provide both basic and applied researchers a high quality avenue to report their work in ...... this exciting aspect of modern engineering. William R. Hamel, ...... The University of Tennessee, USA ...... Mechatronics and industrial automation are at the heart of technological development and it is vital that researchers have high quality publications in which they ...... can rapidly disseminate their results. IJMA will cover a vital and underrepresented area of research and I am Fax ...... confident that it will have a significant impact on the community, quickly becoming a leading journal in this E-mail ...... field. Darwin G Caldwell,

Italian Institute of Technology, Italy Credit card type and number Ê I am honored to be invited to serve as an Editorial Board ...... member of this new journal. Under the guidance of Editors and Editorial Board, this new journal will be an important forum for the publication of exciting new ...... developments in emerging technological areas. James K. Mills, Expiry date ...... University of Toronto, Canada Three-digit security number The birth of the new journal is a big event in the exciting field of Mechatronics and Automation, and I am very (on the reverse of the credit card) ...... proud to be a part of the strong editorial team that will maintain high quality and standards of all the papers Signature...... published in it. The IJMA is certainly going to be a vital and unique source of information on the latest theoretical Date ...... and practical achievements of the researchers in this multidisciplinary and very essential engineering area. Nazim Mir-Nasiri, Please address subscription orders to the Swinburne University of Technology, Malaysia address overleaf Subscription order form

To find out the subscription rates for Relevant International Journal of Mechatronics and Inderscience Titles Automation (IJMA), International Journal of Advanced Mechatronic please go to Systems www.inderscience.com/ijma or http://www.inderscience.com/subscribe.php International Journal of Applied Systemic Studies (for all Inderscience titles, including IJMA) International Journal of Automation and Control This form may be photocopied or downloaded from International Journal of Bio-Inspired Computation www.inderscience.com/www/ielsubsform2.pdf International Journal of Biomechatronics and Biomedical Robotics Journals may be ordered online from http://inderscience.metapress.com International Journal of Computational Vision and Robotics Please address subscription orders to: International Journal of Computer Applications in Technology lnderscience Enterprises Ltd. (Order Dept.) World Trade Centre Building II International Journal of Industrial Electronics 29 Route de Pre-Bois, and Drives Case Postale 856, International Journal of Instrumentation CH-1215 Genève 15, Technology SWITZERLAND International Journal of Intelligent Systems Technologies and Applications For rush orders please: International Journal of Mechatronics and fax: +44 1234 240515 or Manufacturing Systems E-mail: [email protected] International Journal of Modelling, Please enter the following regular Identification and Control subscriptions to IJMA International Journal of Sensor Networks International Journal of System Control and

...... subscriptions (Print or Online) Ê Information Processing [delete as necessary] International Journal of Systems, Control and Communications ...... subscriptions (Print and Online) These titles are part of a unique profile on: ELECTRONIC SYSTEMS, CONTROL AND ARTIFICIAL INTELLIGENCE Total cost ......

For information about these titles and the substantial portfolio of journals developed by Inderscience, please visit the website Methods of payment overleaf at: www.inderscience.com Notes for Authors and Submission of Papers

Submitted papers should not have been previously published or be currently under consideration for publication elsewhere.

All papers are refereed through a double-blind process. A guide for authors and other relevant information for submitting papers are available on the Submission of Papers section of the Inderscience website: please go to

Author Guidelines (www.inderscience.com/guidelines)

To submit a paper, please go to

Submission of Papers (http://www.inderscience.com/papers)

For queries concerning the journal, please contact:

Editor-in-Chief Prof. Shuxiang Guo Kagawa University Department of Intelligent Mechanical Systems Engineering, 2217-20 Shinmachi, Hayashi-cho, Takamatsu, 761-0396, Japan E-mail: [email protected]

With a copy to: Editorial Office, E-mail: [email protected]

(Please include in your email the title of the Journal)

ISSN: 2045-1059 (Print), ISSN: 2045-1067 (Online)

IEEE ICMA 2020 International Conference

本会议由中国吉林省华邦会议展览有限公司支持。

This conference is supported by Jilin Huabang Conference and Exhibition Co., Ltd.