Solving the Combined Modal Split and Traffic Assignment Model with Captive Mode Users and Overlapping Routes

摘要

We developed a path-based partial linearization algorithm for solving the combined modal split and traffic assignment (CMSTA) problem with captive mode users and overlapping routes. The dogit model was adopted to model captive mode users, while the path-size logit (PSL) model was used to handle the route-overlapping problem. A convex mathematical programming formulation was used in this special CMSTA problem, which combines the dogit mode choice and the PSL route choice to form a combined dogit-PSL model. To improve the efficiency of the path-based partial linearization algorithm, a self-regulated averaging line search strategy was incorporated to minimize computational efforts while ensuring convergence. Using a real city network in Seoul, Korea, numerical results were used to examine the efficiency of the proposed algorithm and the effectiveness of how the combined dogit-PSL model handled captive mode users and route overlapping in a multimodal transportation network.