有限理性条件下多层次轨道交通客流拥堵传播机理研究OA北大核心CSTPCDEI

The multi-level rail transit system achieves flexible allocation and complementation of transportation capacity by integrating different modes of rail transit.However,it also increases the uncertainty of passenger flow congestion propagation,which is prone to cross-level propagation of passenger flow congestion.Based on the analysis of passengers'bounded rational choice behavior,the generalized travel time account and congestion degree account were established by introducing the mental accounting theory,and the comprehensive transaction utility of passengers'travel path was described by using the improved value function and decision weight function.By considering the dual factors of travel mode and travel route,a passenger travel mode choice prediction model based on Nested Logit-Mental Accounting-Cumulative Prospect Theory(NL-MA-CPT)was established.On this basis,by introducing the Upgraded Susceptible-Infected-Recovered(USIR)model,a multi-level rail transit passenger flow congestion propagation model considering passenger choice behavior was proposed,and the propagation rate and recovery rate under bounded rationality were characterized.Finally,the Beijing's multi-level rail transit was taken as an example.The effectiveness of the model was verified,and the impact mechanism of relevant key factors on congestion propagation was revealed through sensitivity analysis.The results show that the congestion propagation of passenger flow in multi-level rail transit network has significant volatility and convergence characteristics,and the propagation rate and recovery rate directly affect the intensity and influence range of congestion propagation.With the increase of network layers,the bounded rationality of passenger travel choice decision has a more significant impact on congestion propagation.This study can provide a theoretical reference for operating enterprises to formulate passenger flow congestion control strategies.