Robustness analysis of bus-rail multilayer network based on dynamic passenger flow

In recent years, as public transportation has become the primary mode of travel for urban residents, the passenger load on transit networks has continued to increase. However, most existing studies on cascading failures in transportation networks are based on static topological structures, neglecting the temporal dynamics of passenger flow and their impact on node load distribution and failure propagation paths. Additionally, traditional models often regard passengers as “passive agents†who redistribute uniformly after disruptions, lacking a realistic depiction of travel preferences and behavioral decision-making. This oversimplification fails to capture the nonlinear transfer characteristics of passenger responses under real-world disruption scenarios, limiting the scientific accuracy of robustness assessments. To address these limitations, this study proposes a passenger flow redistribution model that incorporates metro travel preferences and introduces a flow-based edge failure (FE) metric to characterize network degradation under cascading impacts. By integrating mode choice behavior with dynamic inter-node flow transfer rules, a behaviorally responsive cascading failure simulation framework is constructed, offering a new perspective for quantifying the robustness of bus–metro coupled transit networks. Based on data from Chengdu’s bus and metro systems, simulation experiments are conducted for typical time periods on both holidays and weekdays. Results show that during all periods on holidays and the two morning peak periods on weekdays, network robustness is significantly enhanced when passengers prefer metro travel. In contrast, during the two hours following the morning peak on weekdays, higher robustness is observed in the absence of such preferences.