Optimization for operation scheme of express and local trains in suburban rail transit lines based on station classification and bi-level programming

Abstract

In this study, we investigate the train operation scheme in suburban transit lines by considering 1) the station classification and 2) passenger flow assignment. We first classify stations according to the passenger flow. By setting the principle that the express train should stop at the first-level station and skip the third-level station, the solution space of train stop planning is reduced from all stations to the second-level stations. Next, we construct a bi-level programming model for the train optimization scheme under the express/local stopping mode (ELM). The upper level optimizes the operation scheme with the constraints of average vehicle capacity rate, and the reduction of passengers traveling cost (PTC) that compared with traditional (all) stopping mode (TSM), which aims at minimizing agency operations cost (AOC) and PTC; the lower level fully considers passengers route choice behavior and constructs the passenger flow allocation model under the proposed express/local train operation scheme. Owing to the NP-hardness of the studied problem, we propose a heuristic algorithm that combined the genetic-simulated-annealing (G-SA) with the method of successive averages (MSA) that can efficiently solve the proposed bi-level programming model. To evaluate the performance of the proposed algorithm, we present numerical experiments conducted on a real-world case study, based on Shanghai Metro Line 16. The computational results quantify the benefits of considering the combination of station classification and passenger flow assignment in the train operation scheme. Furthermore, we also performed a sensitivity analysis on the weights of AOC and PTC. This analysis provides insights to railway managers on how to set key parameters when applying the proposed formulations and solution methodology in practice.