A Target-oriented Bi-attribute User Equilibrium Model With Travelers’ Perception Errors

In the realistic traffic network,travelers’ route choice behavior can be affected by many factors,such as travel time and travel cost.Meanwhile,due to the complex environment and traveler’s limited cognitive ability,there exist perception errors during their route choice process.Research has also shown that travelers’ route choice behavior can show the behavioral characteristics.Under this background,this thesis adopts a target-oriented analysis method to model travelers’ behavioral characteristics,and at the same time considers the existence of travelers’ perception errors.Based on this,this thesis discusses travelers’ route choice behavior considering the impact of travel time and travel cost.The detailed contents are as follows:First of all,for the two attributes of travel time and travel cost,targets will be specified by travelers during the route choice process.Based on this characteristic,this thesis proposes a target-oriented bi-attribute route utility model with perception errors.There are two typical characteristics of the route utility model proposed in this thesis.One is to use Gaussian copula to capture the stochastic correlation between the perceived travel time and perceived travel cost.In this copula-based framework,the perceived travel time and the perceived travel cost may follow normal distribution in an additive form,lognormal distribution in the multiplicative form,or any other proper distribution.The second is that the interaction between the targets,i.e.,perfect complementary relationship and imperfect complementary relationship,is captured.Travelers are more willing to achieve both targets in the route choice process in perfect complementary relationship.Secondly,based on the target-oriented bi-attribute route utility model with perception errors,the target-oriented bi-attribute user equilibrium with perception errors is further proposed.At this equilibrium,no traveler can increase the route utility by unilaterally changing the route.The user equilibrium proposed in this thesis extends the scope of Wardrop user equilibrium.In the equilibrium model,the targets of route travel time and travel cost are determined endogenously by the traveler’s confidence level,and the confidence level is given exogenously.This thesis formulates this equilibrium model problem as a variational inequality problem,and discusses the existence and uniqueness of its solution.Two related extensions based on the model are also related.One is to discuss target-oriented bi-attribute user equilibrium without perception errors,and the other one is to discuss the multi-class target-oriented bi-attribute user equilibrium with perception errors.Finally,this paper uses the method of successive average to solve the model on the Braess traffic network and the Nguyen-Dupuis traffic network.Based on the equilibrium results,combining travelers’ confidence levels,time-to-cost utility ratio and complementarity ratio,we analytically prove and numerically demonstrate the impact of target on the equilibrium results.The findings of this thesis can provide some new insights into traveler’s route choice behaviors,and the behavioral insights can further provide some implications for the policy design and implementation,e.g.,the congestion pricing scheme.