Route Diversion Probability Model Based on Guidance Utility and Its Application

Abstract

Variable Message Signs (VMS), a collective route guidance information provision, are widely used in urban traffic in most metropolitan areas in China, which actually affect drivers’ route choice behavior. In this paper, a disaggregate choice analysis method is used to describe changes in drivers’ route choice behavior. Logit model, based on guidance utility, is applied to determine probability of route diversion, in which, not only traffic condition information characteristics are included, such as survey time, congestion level, but also drivers’ characteristics are taken into account, such as gender, age, driving experience. A case study is presented by taking drivers’ responses to VMS on Shanghai out-ring road, where a lot of stated preference data and revealed preference data is surveyed. Parameters of the above logit model are determined using those stated preference (SP) and revealed preference (RP) data. The result shows congestion level and age are important factors that affect route diversion probability.