Two new methods for solving the path‐based stochastic user equilibrium problem

Abstract

In this paper, we present two new methods for the path-based logit stochastic user equilibrium problem, and investigate their convergence properties. First, a two level partial linearization method is proposed. Second, a dual method is developed. Both of these two methods use second order approximation of the objective function. Our novel methods are compared to Damberg's partial linearization method (Damberg, 1996), which is known to be one of the best performing methods. Numerical results on the Sioux Falls and Winnipeg networks show that, if properly scaled, our new methods can significantly improve the performance of Damberg’s method.