Departure Time Optimization of Last Trains in Subway Networks: Mean-Variance Model and GSA Algorithm

Abstract

Last-train timetable coordination is extremely complex because a number of transfer directions involve in the subway network. In this paper, transfer redundant time (TRT) and transfer binary variables (TBV) that affect transfer results are considered in the Markowitz mean-variance model. By adjusting running time and dwelling time, the model creates a high-quality timetable that greatly improves the efficiency of transferring passengers. Furthermore, a genetic simulated annealing (GSA) algorithm is developed to solve this problem in the Beijing subway network, which consists of 14 lines, 17 transfer stations, and 42 key directions. The present model increases the number of successful connections by 40.0% and reduces the amount of just-missed connections by 83.3%, respectively. In addition, the mean-variance model significantly improves the subway network accessibility compared with the current practice of the last-train timetable.