Research On Risk-averse User Equilibrium Based On Target-oriented Analysis In Stochastic Traffic Networks

Traveler’s route choice behavior is an important research direction in traffic field.The research on traveler’s route choice can provide reference for the improvement of the traffic system and help to effectively alleviate traffic congestion.Empirical researches show that travelers show risk aversion in the process of route choice.At the same time,empirical researches also show that travelers show behavioral characteristics and pay attention to the influence of acceptable travel time on their route choice.Under this background,this paper adopts the target-oriented analysis method to model traveler’s behavioral characteristics,and at the same time,considering the existence of stochasticity on the supply side of the traffic network,traveler’s route choice behavior under the influence of stochastic travel time has been discussed.The specific contents are as follows:First,this paper model the above two situations concurrently with a target-oriented method,i.e.,a target-oriented method is used to model risk-averse travelers’ acceptable travel time during their route choice.Given target for travel time,travel time that is no larger than this target is termed as acceptable to the travelers.In this case,this target is called has been achieved,and the target achievement can bring utility to the travelers.With this method,two kinds of route choice models are proposed.In the first one,only target achievement is considered,termed as achievement-only route utility model;While in the second one,possibilities that this target is not achieved are additionally considered,termed as failure-aware route utility model.Furthermore,the long-run effect of the proposed route choice models also has been studied,and present the achievement-only and failure-aware user equilibrium.In the proposed user equilibrium,an effective target is presented to model risk-averse travelers’ acceptable travel time.With this target,the model of risk-averse user equilibrium is proved theoretically through analysis.Moreover,a column-generation-based method of successive average is proposed to solve the proposed model on the large-scale traffic network.Finally,numerical examples are conducted to demonstrate the proposed proposition and illustrate the effectiveness of the solution algorithm.The research conclusion can provide some reference for traffic participants and traffic managers,and provides a theoretical basis for alleviating traffic congestion.At the same time,the research of this paper enriches the research content and research ideas of route choice behavior and expands the analysis scope of user equilibrium.