An empirical study of simulation-based dynamic traffic assignment procedures

Under Intelligent Transportation Systems (ITS), real-time operations of traffic management measures depend on long-term planning results, such as the origin-destination (OD) trip distribution; however, results from current planning procedures are unable to provide fundamental data for dynamic analysis. In order to capture dynamic traffic characteristics, transportation planning models should play an important role to integrate basic data with real-time traffic management and control. In this paper, a heuristic algorithm is proposed to establish the linkage between daily OD trips and dynamic traffic assignment (DTA) procedures; thus results from transportation planning projects, in terms of daily OD trips, can be extended to estimate time-dependent OD trips. Field data from Taiwan are collected and applied in the calibration and validation processes. Dynamic Network Assignment-Simulation Model for Advanced Road Telematics (DYNASMART-P), a simulation-based DTA model, is applied to generate timedependent flows. The results from the validation process show high agreement between actual flows from vehicle detectors (VDs) and simulated flows from DYNAMSART-P.