Research on Traffic Assignment Model and Algorithm Based on Cumulative Prospect Theory under Uncertain Factors

Abstract

Current traffic assignment theories and methods are studied. To address the limitations of existing models and algorithms in solving the traffic assignment problems under uncertain supply and demand, a traffic assignment model based on the cumulative prospect theory is proposed. Firstly, the problem of uncertain supply and demand of the traffic network is described, and an improved traffic impedance function of road sections is proposed. Then, the reference point of generalized cost and the reference point of the dynamic section risk are set up, and the comprehensive cumulative prospect value function considering the preference coefficient of the reference point is given. A traffic assignment model based on the cumulative prospect theory is constructed, and an isolation niche genetic simulated annealing algorithm is designed to solve the model. Finally, with the highway traffic network in the Sichuan-Tibet region as an example, the process of network traffic equilibrium based on the cumulative prospect theory under the fixed and changing network structure is studied respectively, and parameter sensitivity and algorithm comparison are analyzed. The results show that the proposed traffic assignment model based on the cumulative prospect theory provides a good idea for solving the traffic assignment problem under uncertain supply and demand, and is of theoretical significance and application value.