Research On Passengers' Time-space Travel Route Estimation Method Of Urban Rail Transit Based On AFC Data

With the increasing complexity of urban rail transit network in China,the trend of large passenger flow becomes more obvious and passenger travel choice tends to diversify.Reasonable description of passenger physical route choice and train selection behavior in the rail transit network is conducive to accurately grasping the distribution law and characteristics of passenger flow in the rail transit network,and provides a scientific theoretical basis for the operation management departments to formulate the operation plan rationally.At the same time,the AFC system of urban rail transit network records a large number of passengers' travel information,which provides real and reliable data support for the study of individual passenger travel behavior in urban rail transit network.Therefore,taking passengers as the research object,this paper proposes a Bayesian-based space-time travel route estimation model to estimate passengers' travel route in the rail transit network in time and space,and to reproduce the physical routes and trains chosen by passengers.Firstly,this paper analyses the characteristics of passenger travel behavior in urban rail transit,and divides travel time into entry time,transfer time,exit time and in-vehicle time.Secondly,it defines the passenger's space-time travel route,searches the effective physical route set based on Dijkstra algorithm and branch-and-bound idea,and proposes an arc time constrained travel route search algorithm to generate the passenger's effective space-time travel route set.Then,based on the characteristics of the effective physical path number,the number of passenger schemes and the number of transfers,the passenger itineraries are classified.Based on Bayesian theory and frequency statistics,the probability distributions of exit time,entry time and transfer time are estimated.Based on the time probability distribution,the matching probability between passengers and alternative space-time travel routes is estimated,and the passengers in urban rail transit are obtained.Historic space-time travel route in rail transit network.Finally,taking Beijing urban rail transit system as an example,the actual AFC data and train operation data of a working day are selected for the application of the model,and the apace-time travel route selection probability of the selected typical passengers is estimated.This study supports the reappearance of individual historical trajectory in urban rail transit network,which is conducive to grasping the rules of passenger travel choice and the characteristics of passenger flow distribution,and provides scientific support for the operation organization of urban rail transit.There are 18 pictures,8 tables and 59 references.