Measuring and enhancing the connectivity reliability of a rail transit network

Abstract

A model for estimating the connectivity reliability (CR) of a rail transit network (RTN) is proposed that considers the passengers’ travel behavior. Passengers choose acceptable paths whose trip times are below the passengers’ acceptable trip time. An origin-destination station (OD) pair’s CR is defined as the probability that at least one acceptable path is connected between that OD pair. The RTN’s CR is defined as the average value of CR for each passenger on the RTN. A model is proposed to maximize an RTN’s CR by adding trains, subject to constraints on operational cost, allowable track capacity and available vehicles on each line. The model is solved with a multi-population genetic algorithm (MPGA). The model application in Chengdu’s RTN shows that adding trains corresponding to the optimized solution with an operational cost constraint, not only increases the net benefit of RTN operations, but also enhances the RTN’s CR.