A Multi-Mode Variable Demand Traffic Assignment Problem for the Transportation Network Investment

In this paper, the authors develop a multi-mode variable demand traffic assignment problem required in the transportation network investment decision. Transportation network investment scenario analyzes the future construction plans on the road and rail networks using estimated demand and mode-choice. The key output from this analysis is the user equilibrium travel time for each mode. Modes considered in this paper are auto, bus, truck, passenger and cargo rails. The authors develop a multi-mode variable demand traffic assignment model based on the inverse demand function and excess flows. For the excess flow model, the authors provide the optimality condition and for the inverse demand function model, they develop a partial linearization algorithm. Both approaches are tested using a large scale network consisting of 2,192 nodes, 5,735 links, and 17,424 origin-destination pairs and the authors show implementation details enhancing the stability of the algorithm.