Work

Integrated metro train timetable and rolling stock rescheduling with passenger rerouting under multi-location disturbances

Year: 2026

Type: article

Abstract: The expanding scale of infrastructure and rising passenger demand in metro systems increase the frequency of stochastic disturbances, significantly impacting metro train operations. This is particularly evident in multi-location disturbances (MLD), often occurring during extreme weather events, equipment failures, and unexpected passenger demand. Compared to the single-location disturbance scenario, the decision-making process for train rescheduling in the MLD scenario is inherently more complex, posing challenges in mitigating passenger congestion at stations and ensuring the feasibility of rolling stock circulation, including rolling stock turn-back and deadheading operations. To enhance the quality of rescheduling decisions under the MLD scenario within limited computational times, this study investigates a range of comprehensive rescheduling strategies, including holding, retiming, train cancellation, short-turning, skip-stopping, and rolling stock reassignment. Recognizing the necessity for passenger rerouting due to the application of short-turning and skip-stopping strategies, we develop a mixed integer linear programming model for metro train timetable and rolling stock rescheduling, aiming to minimize the weighted sum of four objectives: deviations with the planned train timetable, the number of backup rolling stocks used, additional average passenger travel times, and passenger gathering risks. We introduce various auxiliary variables to capture the coupling relationship between the application of rescheduling strategies while employing a discretization mechanism to facilitate the linearization of passenger rerouting constraints. To effectively address the problem and improve the quality of the train rescheduling solution, we develop a novel hybrid simulation-optimization (HSO) algorithm based on the rescheduling strategies mentioned above. Finally, real-life case studies from Chongqing Rail Transit Line 5 demonstrate that the proposed algorithm delivers superior rescheduling solutions efficiently, with an average improvement in the objective value of 47.59% over the two-stage method, 20.83% over a standard local search algorithm, and 9.57% over the HSO algorithm without implementing the tabu search mechanism.

Source: Transportation Research Part B Methodological

DOI: 10.1016/j.trb.2026.103603

Authors: Shan Jiang, Jiateng Yin, Yongxiang Zhang, Andrea D’Ariano, Xinyue Hu, Qiyuan Peng

Institutions: Beijing Jiaotong University, Southwest Jiaotong University, Roma Tre University

Language: English

Topic: Railway Systems and Energy Efficiency

Subfield: Industrial and Manufacturing Engineering

Field: Engineering

Domain: Physical Sciences

SDGs: Sustainable cities and communities, Industry, innovation and infrastructure

Open Access status: hybrid

Estimated APC paid: $3,810