Individualized Gap-Based Convergence in an Agent-Based Dynamic Traffic Assignment Model Using an Information Mixing Approach for Time-Dependent Travel Times

摘要: This study proposes an individualized convergence approach for the agent-based dynamic traffic assignment problem. The name agent-based comes from the fact that each traveler is routed individually from their specific origin to their specific destination at a specific departure time. The approach is gap-based two-folds. Firstly, the re-assignment decision is based on the gap between the routed and experienced travel time from the previous iteration, which is an existing method in the literature. Secondly, the historical time-dependent and prevailing traffic conditions are averaged using the weight calculated by a modified two-parameter Weibull survival function. This weight is individualized based on the relative gap of the traveler from the previous iteration, as well as the iteration number. This information mixing approach is a novel contribution of this study. The methodology is tested on a medium-scale network of Bloomington, IL in the United States of America. The algorithm converges after only two iterations, which is very promising especially for large-scale networks where the computational time of a single iteration can be very high.