Research Article Minimizing Metro Transfer Waiting Time with AFCS Data Using Simulated Annealing with Parallel Computing
Total Page：16
File Type：pdf, Size：1020Kb
Hindawi Journal of Advanced Transportation Volume 2018, Article ID 4218625, 17 pages https://doi.org/10.1155/2018/4218625 Research Article Minimizing Metro Transfer Waiting Time with AFCS Data Using Simulated Annealing with Parallel Computing Xiaobo Liu ,1,2 Minghua Huang ,1 Hezhou Qu ,1,2 and Steven Chien 3 1 School of Transportation and Logistics, Southwest Jiaotong University, Chengdu, Sichuan 610000, China 2National Engineering Laboratory of Integrated Transportation Big Data Application Technology, Chengdu, Sichuan 610000, China 3John A. Reif, Jr. Department of Civil and Environmental Engineering, New Jersey Institute of Technology, Newark, NJ 07102-1982, USA Correspondence should be addressed to Hezhou Qu; [email protected] Received 16 April 2018; Revised 10 September 2018; Accepted 18 September 2018; Published 4 October 2018 Academic Editor: Luigi Dell’Olio Copyright © 2018 Xiaobo Liu et al. Tis is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Coordinating train arrivals at transfer stations by altering their departure times can reduce transfer waiting time (TWT) and improve level of service. Tis paper develops a method to optimize train departure times from terminals that minimizes total TWT for an urban rail network with many transfer stations. To maintain service capacity and avoid operational complexity, dispatching headway is fxed. An integrated Simulated Annealing with parallel computing approach is applied to perform the optimization. To demonstrate model applicability and performance, the Shenzhen metro network is applied, where passenger fows (i.e., entry, transfer, and exit) at stations are approximated with the automatic fare collection system (AFCS) data. Results show that the total TWT can be signifcantly reduced. 1. Introduction Oyster Card in London, etc.). With the AFCS data in this study, the temporal and spatial passenger entry/exit distri- For large-scale metro networks, large numbers of passengers butions and transfer volumes at stations can be determined. arrive simultaneously at transfer stations, especially during We aim to seek a quick and easy-to-implement method peak periods. Transfer time incurred by passengers has been to efectively reduce TWT without increasing operator’s considered as an index of service quality. Synchronizing cost and operational complexity. To this end, we alter the vehicle arrival times to facilitate efective timed transfer is train departure times from terminals but fx the dispatching desirable. headways. Te research problem is combinatorial, which Transfer time is defned in this study as the elapsed time is difcult to solve with classic optimization methods. An between a passenger alighting from a “delivery” train to integrated Simulated Annealing (SA) with parallel computing boarding a “pickup” train, which consists of walking time (PC) approach is developed to search for the solution that between platforms and transfer waiting time (TWT) for the minimizes TWT. “pickup” train. Walking time is dependent on the layout of Te remainder of this paper is structured as follows. the station and walking speed, while TWT is afected by the Section 2 describes previous research and practices related arrival/departure times of the trains and walking time. To to this study. Te development of the proposed model and minimize TWT for a large metro network, the relationship solution method to solve the study problem are discussed between demand (i.e., spatiotemporal transfer passenger in Sections 3 and 4, respectively. In Section 5, the Shenzhen distribution) and supply (i.e., network confguration, service Metro in China is employed as a case study to demonstrate frequency, and departure time) must be carefully formulated. the efectiveness of the solution algorithm and the benefts of Automatic fare collection systems (AFCS) have been the optimized solution. Finally, the paper concludes with a applied in many metro systems (i.e., Metro Card in New York, summary of fndings and future research needs. 2 Journal of Advanced Transportation 2. Literature Review minimized total transfer wait time. Using GA and the Frank- Wolfe algorithm, Xiong et al. [28] synchronized arrivals of Coordinating train arrivals to facilitate transfer activities in community shuttles linking with a metro service, which a transit network ofers an efective way to reduce transfer yielded the minimum total cost. Due to the absence of the time. Transfer time is an important indicator representing AFCS data, the distribution of passenger arrivals at stations in the service quality and the efciency of public transit systems most previous studies discussed above was either simplifed [1–3]. Previous studies have focused on minimizing total or assumed. Few studies have applied the AFCS data for waiting time [4–8] or maximizing the number of synchro- analyzing passenger demand and route choices [29–31] and nized vehicle arrivals [9, 10] at transfer nodes. Recently, optimizing timetable [32]. several studies focused on the frst and last trains timetabling Te primary contribution of this study is to collect, optimization, with objectives such as maximizing passenger process, and generate the spatiotemporal origin-destination transfer connection headways [11], minimizing the connec- (OD) demand with the AFCS data, approximate transfer tion time between frst trains [12], incorporating bilevel volumes at stations, and apply that to minimize total TWT objectives in which the upper level maximizes social beneft for a large metro network with many transfer stations. Opti- (the number of passengers transferring to the last trains mizing such a coordinated transfer problem is challenging successfully) with minimum total subsidy and the lower because it is combinatorial and difcult to solve with classic level minimizes revenue loss for the operating companies optimization methods. A method that integrates SA with [13], and maximizing the total number of successful transfer PC (SAPC) is developed and implemented to search for the passengers and minimizing waiting times for rail-to-bus optimal solution. passengers [14]. Several studies optimized coordination of transfers at a 3. Methodology single hub station. Feng et al. [15] developed an optimization model for the arrival and departure times of connecting Te mathematical model is formulated and discussed in this trains, which minimized average waiting time. Liu et al. [16] section, and the objective is to minimize total TWT. To developed a multiobjective optimization model to minimize maintain service capacity and avoid operational complexity, the total transfer waiting time, train operating cost, and the preplanned dispatching headways are fxed. Te decision fuctuation of departure interval with given transfer volumes. variables therefore consist of train departure times from the Te optimized results were found using a genetic algorithm beginning terminals of all lines in the study network. Te (GA). Chen and Wang [17] minimized passenger waiting time development of the proposed model is discussed next, and via synchronizing train arrivals and departures and revealed the variables used to formulate the model are summarized in the relationships between three infuencing factors (transfer Table 1. forms, walking distance in the station, and congestion degree of passenger fow) and transfer time. 3.1. A General Metro Network. A general metro network Optimizing coordination among multiple transfer sta- consists of many routes defned as � and a set of stations defned as . A route is represented by two unidirectional tions was conducted by previous studies. Assuming uniform passenger arrivals, Wong et al. [18] presented a mix-integer lines (i.e., outbound and inbound). Line belongs to a set of lines denoted as �. Each line consists of a set of stations programming model to optimize nonperiodic timetable. A ∈ heuristic approach was employed to minimize the waiting �,and � . Each station is given a unique station ID. For example, station ID of line 1 begins with 1 and ends at time by justifying the train running time between stations, +1 dwell times at stations, and departure and turnaround times 1, and then the IDs of stations on line 2 start from 1 and end at 2.Tus,stationsoflinel ( = 2,3,...,) are at terminals. Fang et al. [19] minimized total waiting time +1 yielded by the optimized vehicle arrival and departure times labeled from �-1 through �.Forlinel, the hourly service frequency is denoted as �, and a train is indexed by m (= at transfer stations. Aksu and Akyol [20] optimized integer- 1,2,..., ratio headways to minimize total cost. Considering random �). To formulate TWT incurred by passengers at a passenger arrivals, Liu et al. [21] minimized total waiting transfer station, a link-node diagram is shown in Figure 1 where transfer station is also labelled as � for line as well time by altering the train departure times found by SA. To � ,� ∈ improve transfer efciency, Wu et al. [22] optimized vehicle as �� for line ( � � ). Transfer station has four IDs departure time, headway, running time, and dwell time using associated with four connecting lines. GA. To formulate the proposed model, the following assump- tions are made and subsequently explained: Considering probabilistic vehicle travel time, Chowdhury and Chien [23, 24] synchronized vehicle arrivals for an (1) Te platform-to-platform walking time at a transfer