Transfer Volume Forecasting Method for the Metro in Networking Conditions

Abstract

Transfer volume forecasting is the basis of transfer station design, passenger flow organization and emergency scheme formulation. Along with rapid development of Metro networking operation in recent years, transfer volume forecasting for the Metro system in networking conditions becomes an important issue to be solved. In this paper, the transfer volume forecasting method was proposed on the basis of route choice analysis and passenger flow distribution. Taking into consideration the variables of level of service such as the ride time, number of transfers, transfer time and congestion degree and introducing in the index of angular cost value to describe the influence of network structures and route directions on route choices, the route choice model was established to analyze passengers' route choice preferences. Through equivalent ride time conversion of the above influential factors and analysis on the degree of passenger tolerance to travel time, the effective route set of all OD (Origin-Destination) pairs was constructed with the K-shortest paths algorithm. Then stochastic assignment was carried out with the MSA (method of successive averages) algorithm under constraint of the first and last train and the statistical method was put forward to calculate the transfer volumes of all transfer stations within a specified time interval. Finally, the transfer volumes of the Beijing Subway were estimated with the proposed transfer volume forecasting method. The results show that the angular cost value significantly influences the route choice behavior and the proposed method can provide good reliability and accuracy.