MATEC Web of Conferences 139, 00043 (2017) DOI: 10.1051/matecconf/201713900043 ICMITE 2017 Heliostat attitude angle detection method based on BP neural network LIU Guangyu1,*,CAI Zhongkun1 1School of Automation Engineering, Hangzhou Dianzi University, Hangzhou 310018, China Abstract: Traditional fossil fuels have dried up, global warming and sustained economic development have led to the rapid growth of clean energy resources. Tower thermal power generation has attracted much attention due to its ability to generate electricity during the night. The traditional tower thermal power generation adopts open-loop control which requires very high mechanical accuracy. In the operation of power station and there may be a settlement, wind load or other factors make the heliostat skew phenomenon. It will eventually lead to a decline in power generation efficiency. Thus, we propose a closed-loop feedback control method based on machine vision and optical reflection principle based on the method of using the correction of heliostat spot acquisition board. To identify the spot and the ellipse fitting method for spot feature extraction using image processing technology, we propose a heliostat to determine the characteristics of the corresponding spot mapping the attitude angle method based on BP neural network. Thus we can provide direct feedback control of heliostat errors. The new method can effectively increase the heliostat tower power generation efficiency and also can make the tower heliostat thermal power generation cost reduced with the popularization and application of significance. precision, resulting in high construction costs . The literature[5] used the camera as the day to achieve 1 Introduction accurate tracking sun trajectory tracking, but because of As a clean and renewable energy, solar energy attracts the weather, the application will have certain limitations more and more attention. Solar power is divided into two and difficulties in practical application; Literature[6] categories: photo-voltaic and photo-thermal power proposes the use of Whiteboard Based correction generation. And the technology faces enormous pressure methods. But because of the way based on spot to reduce the cost of electricity as soon as possible[1]. correction, not directly to optimize the layout of the Because of the thermal power generation tower has heliostat field to provide effective data. special advantages, has attracted much attention, but The above mentioned in the literature are not because the control precision, heliostat operation stability, involved with heliostat attitude, but in practical safety and reliability and the construction cost is limited. application, heliostat attitude angle correction for In order to reduce the cost of tower thermal power heliostat reflection is very important[18]. In the tower generation[2,8,9,12]. heliostat thermal power generation, because of the Tower power by elevation and azimuth tracking heliostat tower distance target center is far away, it will control of heliostat. The heliostat tracking control cause deviation of the heliostat spot cannot be accurately method for main program control and sensor control and reflected to the target position, if the deviation is too sensor program, hybrid control in three ways[5]; in the large, it will even cause the spot projected on the target practical application, mainly adopts open-loop control position. based on program control. The procedure control and the This paper presents a method of heliostat attitude use of sensor data is gradually become the focus of the angle from the spot of direct mapping. In many heliostat [3,4,13]. applications of new energy systems, such as fault Heliostat traditional open loop control directly using diagnosis, neural network method can be used as a the clock to control the rotation angle of the heliostat, powerful tool for processing, and has a very good very high requirements for heliostat support mechanical effect[14,19,20].The complex neural network structure can approximate any nonlinear function, and can solve * Corresponding author: [email protected] © The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0 (http://creativecommons.org/licenses/by/4.0/). MATEC Web of Conferences 139, 00043 (2017) DOI: 10.1051/matecconf/201713900043 ICMITE 2017 the complex nonlinear mapping problem very well. This The characteristic value of the ellipse is equivalent to paper uses BP neural network mining implicit mapping the feature of the reflected facula. Thus, the feature relationship between spot feature and heliostat attitude extraction of the spot is completed. Based on the data angle. Based on the small biaxial heliostat tracking source for the identification of heliostat attitude angle. platform for data acquisition and verification experiment, Since the establishment of the model directly from the method can accurately identify the heliostat spot with the heliostat reflection spot a heliostat precise attitude the characteristics of the corresponding attitude angle. angle is difficult. Considering only the heliostat elevation angle and azimuth axis moving, so under the given conditions there will be corresponding with the only 2 Problem description spot[11]. There is a map that reflects the pattern characteristics of heliostat attitude angle. In the tower heliostat thermal power generation, we need In view of the above situation, this paper proposes a the heliostat reflection light reflected to the specified data driven model identification method based on BP position of the target tower to achieve maximum heat neural network. To solve this problem, this paper use the collecting efficiency and Maximize output [6]. In method of image processing based on the spot processing engineering applications, due to the installation error, target spot and find the corresponding elliptic equation mechanical wear and other irresistible factors will make by the method of ellipse fitting. five features: Based on the heliostat cannot be accurately reflected to the target area. Therefore, in practice, is the attitude angle of the ellipse long axis a , short axis b , inclination of the long heliostat need correction, this paper proposes a axis , the coordinates of the center of xc and yc . In correction method for high precision heliostat attitude the specific time and location and coordinates, this paper angle based on BP neural network. presents a mathematical mapping for two spot attitude According to the physical model of heliostat . characteristics and heliostat angle between a and z reflection [6,7]. At the specified coordinates by the The following mapping: relevant input parameters, we can get the heliostat azimuth and elevation angle. The reflected light spot is fa ():((),() t at bt , xca(t ),ytc ( ), ( t )) ()t (7) projected to the target area, and then the original image fz ():((),(), t at bt xcz(),t ytc (),() t) ()t (8) data is collected. In order to use the data effectively and get the best spot, we need to preprocess the original data [10,15,16,17].After the image processing, the smaller 3 Heliostat angle detection method size of the reflected spot image is obtained. Based on the based on BP neural network least square method, ellipse fitting is used to obtain the general characteristic equation of fitting ellipse, and the algebraic form of conic equation is expressed as formula 3.1 Method design (1) In order to solve the above problems, a flow chart is Ax22 Bxy Cy Dx Ey F 0 (1) designed using the method shown in the diagram. Among The five eigenvalue representations of the ellipse them, the BP neural network model is the main problem can be obtained from the characteristic equation of the solving tool used in this paper. ellipse, that is, the coordinates of the central point of the ellipse, the length of the long axis, the axis of the ellipse and the inclination of the ellipse. Its algebraic expressions, such as formula (2) to (6) BE 2 CD x (2) c 4AC B2 BD 2 AE y (3) c 4AC B2 2F a 2 (4) AC AC B22 () F 2F b 2 (5) 22AC . AC B () F Figure. 1 flow chart of method design 1 B arctan (6) It can be seen from the picture, the mapping between 2 AC the unknown heliostat reflection spot feature and 2 MATEC Web of Conferences 139, 00043 (2017) DOI: 10.1051/matecconf/201713900043 ICMITE 2017 - 3 - the complex nonlinear mapping problem very well. This The characteristic value of the ellipse is equivalent to heliostat attitude angle. The model identification process from the output to the input direction. Table 1 lists the paper uses BP neural network mining implicit mapping the feature of the reflected facula. Thus, the feature mainly includes the following five parts: mathematical symbols used in the model and their relationship between spot feature and heliostat attitude extraction of the spot is completed. Based on the data Step one: model identification using standard BP implications. angle. Based on the small biaxial heliostat tracking source for the identification of heliostat attitude angle. neural network, aiming at the problem of the analysis platform for data acquisition and verification experiment, Since the establishment of the model directly from results, application of this issue to be resolved for the Table 1 ANN Symbol specification the method can accurately identify the heliostat spot with the heliostat reflection spot a heliostat precise attitude mapping relationship between heliostat spot feature and Symbol Implication the characteristics of the corresponding attitude angle. angle is difficult. Considering only the heliostat elevation heliostat attitude angle, the model identification using Learning rate angle and azimuth axis moving, so under the given standard BP neural network. p Sample p conditions there will be corresponding with the only Step two: experimental design, in order to obtain The connection weight between the input 2 Problem description spot[11]. There is a map that reflects the pattern experimental data for training neural networks, we use a wij layer and the hidden layer neuron i through characteristics of heliostat attitude angle.