Research On Passenger Route Choice Behavior Under Disruption Of Urban Rail Transit Based On Cumulative Prospect Theory

With the continuous development of urban rail transit and the sustained increase of passenger flow,the disruption of urban rail transit occurs from time to time,which brings a greatly negative impact on passenger travel and operations management.It is of great significance to accurately grasp passenger route choice behavior and distribution characteristics of passenger flow under the disruption,which is the premise of formulating reasonable and effective emergency measures.Aiming at the disruption of urban rail transit,this paper models passenger route choice behavior and constructs a passenger flow assignment algorithm based on cumulative prospect theory.The main research contents include:(1)Theoretical analysis.The definition and the classification of disruptions are elaborated,and then the research category of this paper is determined as the bidirectional disruption occurring in the partial interval of the line within 30 minutes.Based on the characteristics of passenger route choice behavior under the disruptions,it is proposed that passengers travel decision process under the disruption can be divided into two stages: route choice under the normal and route choice under the disruption and the results of the former are a reference to the latter.(2)Passenger route choice models under the normal and the disruption are constructed respectively.For the normal,taking path properties as utility,a route choice model based on the improved MNL model is constructed.For the disruption,a route choice model combining cumulative prospect theory and NL model is constructed taking the cumulative prospect value of the path as utility.The parameter calibration and validity of the models are verified by a case,and the dynamic passenger route choice behavior under the change of disruption duration is analyzed.The results show that the change of disruption duration has a significant impact on passenger route choice behavior.(3)Passenger flow assignment algorithm is constructed to allocate the passenger flow under the disruption.The effective path search algorithm based on depth first and the reconstruction principle under the disruption are constructed.The dynamic inference method based on individuals is proposed to deduce the real-time location of passenger flow.The classification of passenger flow under the disruption is given and the corresponding assignment principle is determined.Combined with the dynamic inference method and the assignment principle,the passenger flow assignment algorithm under the disruption is constructed to realize the passenger flow assignment in the whole process varying with the duration from the beginning to the end of the disruption.(4)Case analysis.Taking Beijing rail transit network as an instance,the operation data needed for passenger flow assignment are sorted out and processed,and OD passenger flow demand during morning peak hours is obtained through AFC data.The corresponding disruption scenarios are set to realize the assignment of passenger flow under the normal and the disruption of Beijing rail transit during morning peak hours.The results of passenger flow assignment are compared and analyzed from the following dimensions: inbound and outbound flow,section flow,transfer flow,affected flow,the passenger aggregation number at the station,and the degree of impact on the station.The results show that the passenger flow distribution under the disruption has the characteristics of disequilibrium and strong fluctuation,and the impact of the disruption on the passenger flow has the characteristics of propagation,lag and agglomeration.There are 40 figures,25 tables and 78 references in the thesis.