Disruptions can render parts of the critical transportation systems unavailable, forcing both trains and passengers to adapt. This study addresses the integrated rescheduling problem in a high-speed railway network during severe disruptions, focusing on train routing, timetable adjustments, and passenger reassignment. We employ rescheduling strategies that allow disrupted trains to reroute through alternative paths within stations and across the network, utilizing remaining capacity to ensure reliable service for affected passengers. To tackle this issue, we propose a path-based mixed-integer linear programming (MILP) model based on detailed space–time networks, aiming to minimize total train delays and passenger inconvenience caused by disruptions. However, solving this integrated model using the column generation method presents convergence challenges as the problem scale increases. To address these challenges, we introduce a hierarchical solution framework with two main components: (1) a Benders decomposition-based procedure to iteratively capture the interaction between train rescheduling and passenger reassignment, and (2) two column generation procedures to explore promising space–time paths for both trains and passengers. Additionally, a dynamic constraint generation technique is integrated to further accelerate the solution process. Numerical experiments using real-world data from Chinese high-speed railway network validate the effectiveness of the proposed approach. The results show that our method delivers high-quality solutions within an acceptable time frame, efficiently reassigning passengers and rerouting trains during disruptions. Experimental findings also reveal that integrated modeling improves overall efficiency by 17.32% on average compared to sequential modeling. Furthermore, the proposed hierarchical algorithm significantly outperforms traditional column generation methods, reducing computation time by an average of 53.82%.
Xiu, C., Pan, J., D'Ariano, A., Zhan, S., Tessitore, M.L., Peng, Q. (2025). Integrated train rescheduling and passenger reassignment for disrupted high-speed railway networks: A hierarchical Benders decomposition and column generation approach. TRANSPORTATION RESEARCH PART E-LOGISTICS AND TRANSPORTATION REVIEW, 200 [10.1016/j.tre.2025.104177].
Integrated train rescheduling and passenger reassignment for disrupted high-speed railway networks: A hierarchical Benders decomposition and column generation approach
D'Ariano A.;Tessitore M. L.;
2025-01-01
Abstract
Disruptions can render parts of the critical transportation systems unavailable, forcing both trains and passengers to adapt. This study addresses the integrated rescheduling problem in a high-speed railway network during severe disruptions, focusing on train routing, timetable adjustments, and passenger reassignment. We employ rescheduling strategies that allow disrupted trains to reroute through alternative paths within stations and across the network, utilizing remaining capacity to ensure reliable service for affected passengers. To tackle this issue, we propose a path-based mixed-integer linear programming (MILP) model based on detailed space–time networks, aiming to minimize total train delays and passenger inconvenience caused by disruptions. However, solving this integrated model using the column generation method presents convergence challenges as the problem scale increases. To address these challenges, we introduce a hierarchical solution framework with two main components: (1) a Benders decomposition-based procedure to iteratively capture the interaction between train rescheduling and passenger reassignment, and (2) two column generation procedures to explore promising space–time paths for both trains and passengers. Additionally, a dynamic constraint generation technique is integrated to further accelerate the solution process. Numerical experiments using real-world data from Chinese high-speed railway network validate the effectiveness of the proposed approach. The results show that our method delivers high-quality solutions within an acceptable time frame, efficiently reassigning passengers and rerouting trains during disruptions. Experimental findings also reveal that integrated modeling improves overall efficiency by 17.32% on average compared to sequential modeling. Furthermore, the proposed hierarchical algorithm significantly outperforms traditional column generation methods, reducing computation time by an average of 53.82%.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
