A local search algorithm for finding optimal transit routes configuration with elastic demand

This paper proposes a new local search algorithm for finding the optimal configuration of subroutes from a set of candidate transit routes in a transportation network. It is intended to maximize the transit ridership while holding the budget constraint. In each iteration of the algorithm, route segments that are likely to absorb more transit passengers are added to the configuration and less-contributing segments are removed, instead. A path-based model with elastic demand is applied for traffic assignment problem. The algorithm takes advantage of the equilibrium paths information to speed up the calculations for emerging configurations. A numerical experiment on Sioux-Falls network indicates that the proposed algorithm can achieve high-quality solutions at different levels of budget. Also, the run-time and performance of the algorithm are reported over a large problem instance of the Chicago sketch network with 55 artificial candidate routes.