A Multi-Periodic Train Timetable Optimization Model for Suburban Railway Considering Uncertain Passenger Demand

Suburban passenger demand typically exhibits unbalanced distributions, characterized by highly heterogeneous spatial mobility patterns and flow sizes. Designing an efficient, regular, and stable train timetable is critical for long-term use in suburban railways. This study adopts optimization methods to tackle the multi-period train timetable problem, considering both periodic and aperiodic nature to accommodate strongly heterogeneous train services. An MP-TTP model is constructed to minimize travel time and maximize timetable robustness. Additionally, the study introduces a multi-scenario demand input method to address the uncertainty of daily passenger demand. A hybrid particle swarm optimization and genetic algorithm (PSO-GA) is designed. The proposed approach is verified by a real-world case study. The results suggest that the multi-scenario-based train timetable generally outperforms the independent scenario-based method in terms of improving robustness. Furthermore, the multi-periodic train plan can effectively balance efficiency and accessibility while considering regularity.