A Stochastic User Equilibrium Model and Optimal Congestion Pricing with Prospect Theory

In spite of the fact that expected utility theory is a main method to model people's travel choice behavior under uncertainty, the assumption that each decision maker (DM) is completely rational limits its wide application. This paper applies prospect theory, which can appropriately describe DMs’ bounded rational behavior, to route choice behavior analysis and develops an improved stochastic user equilibrium model and finally analyzes optimal congestion pricing. We find that the prospect in prospect theory could be regarded as a special utility in some degree. Based on this result we assume that the prospect of each route is constituted by a fixed term and a random one, and then proposed the concept of perceived prospect. Further assumption is that the random variable follows some certain normal distribution and then a probit model is put forward. Considering the complexity of probit model, Monte Carlo simulation is brought into the solution algorithm of MSA. A numerical example is given to show the impact of uncertainty, including subjective and objective uncertainty that travelers have to deal with. Congestion pricing is also taken into account to express the influence of traffic management policy on traffic flow on a network. At last, an approximately optimal charging standard is proposed through comparing one-link charging and two-links charging. The results prove the effectiveness of congestion pricing. Further research shows that there are 1.91% and 2.71% drops of total travel time and total travel cost respectively when achieving the optimal charging standard.