Abstract
This paper proposes a bi-level programming model to determine the optimal lane configuration for contraflow operations. For the bi-level programming model, the upper level problem aims to minimize the total travel time of a study area by choosing the appropriate number of lanes of candidate links to reserve their travel directions, which is formulated as an integer programming; the lower level is a logit-based stochastic user equilibrium (SUE) traffic assignment model that is able to predict the network flow pattern with respect to the change of network topological structure made by the upper level problem. Furthermore, this paper also proposes a hybrid genetic algorithm embedded with a SUE traffic assignment method for solving the proposed bi-level programming model. The proposed model and solution algorithm are tested using the Sioux Falls network. Results show that the methodology proposed can produce promising results.
