A regional traffic simulation/assignment model for evaluation of transit performance and asset utilization: time dependent multimodal and intermodal assignment models

Abstract

This report presents three approaches to modeling multimodal and intermodal transportation network problems for evaluation of transit performance and asset management. The first approach is an automobile assignment-based multimodal approach, which captures bus movements in the simulator, but assigns only automobile trips to shortest paths under the assumption that the mode split is fixed. For this model, an inner assignment dynamic user equilibrium (IADUE) assignment algorithm was developed to replace the method of successive averages (MSA). Whereas the MSA approach, which has been the state of the art in simulation-based assignment models, assigns vehicles to paths probabilistically, the IADUE approach systematically searches for the equilibrium assignment. Next, a person assignment-based intermodal approach is presented such that mode and path choices are modeled as simultaneous decisions. This approach uses the same multimodal car and bus simulator as in the vehicle assignment approach, and also uses the IADUE algorithm to determine the equilibrium path assignment; however, instead of automobile-only shortest paths, intermodal least cost paths are calculated at each iteration. Third, an integer linear programming formulation of the system optimal intermodal assignment model is presented. Computational results are presented for all three models.