Depth-first Search Algorithm with the Method of Successive Averages for Solving the Logit-based Stochastic User Equilibrium Problem

Abstract

Logit-based stochastic user equilibrium (SUE) traffic assignment plays a vital role in the range of traffic assignment models. The logit-based SUE problem could be resolved by using some efficient algorithms in which the link-based approaches avoiding route enumeration are preferable. To contribute to the system of these algorithms, this paper proposes the new algorithm based on the Depth-first search algorithm with the method of successive averages for solving the logit-based SUE problem. On one hand, the proposed algorithm could keep the advantage of preventing the route variables that reduces computation storage, on the other hand, it is easy to apply in practice and will also reduce the computation time in some cases of testing. Besides, the new algorithm is more applicable in some extensions of the logit-based model. In this study, Kanazawa city urban of Japan application is adopted to demonstrate the effectiveness of the proposed method.