Bayesian network modeling explorations of strategies on reducing perceived transfer time for urban rail transit service improvement in different seasons

Abstract

A new Bayesian Network (BN) learning approach is developed in this work to analyze the effect of different factors on the Perceived Transfer Time (PTT) of the Urban Rail Transit (URT) passengers. It is shown that the newly developed approach is able to build a BN with a satisfactory ability to assess effective strategies on reducing the PTT for the URT service improvement. Moreover, it is found that mainly determined by the weather, the relative environment inside an URT station plays the key role in deciding the impacts of varied factors on the PTT in different seasons. Fully illuminating the transfer passageway and preparing adequate and clear transfer guidance in an URT station are the most important in spring for the reduction of the PTT. Effectively decreasing the passenger flow conflicts in an URT station is the only way to evidently reduce the PTT in summer. Not only fully illuminated and well decorated transfer passageway but also increased train service frequency is indispensable to the decrease of the PTT in autumn. Besides minimizing the passenger flow conflicts, improving both frequency and time coordination of train services is essential for reducing the PTT in winter as well.