Real-time Dynamic O-d Matrix Adjustment Using Simulated and Actual Link Flows in Urban Networks

The present paper investigates the efficiency and robustness of different real-time dynamic origindestination (O-D) matrix adjustment algorithms when implemented in large-scale transportation networks. The proposed algorithms produce time-dependent O-D trip matrices based on the maximumentropy trip departure times, as they are calculated with the use of simulated and actual observed link flows. The implementation of the algorithms, which are coupled with a quasi-dynamic traffic assignment (DTA) model, showed their convergent behavior and their potential for handling realistic urban-scale network problems, in terms of both accuracy and computational time. The algorithm performance was found to be affected by the assumptions underlying the structure of the prior matrix information and the quasi-DTA model, the time interval duration, the nature of observed link flows, the network scale and the link count availability. The relative efficiency of the algorithms was found to depend on the level at which the assigned flows approximate the observed link flows. These results may provide insights on the suitability of each algorithm for diverse application domains, including freeways, small networks and large-scale urban networks, where different quality of O-D information is usually available. Furthermore, other sources of traffic information, in addition to the link counts, could possibly enrich the evaluation of the performance of dynamic matrix adjustment procedures for the case of real-time urban networks. TRB 2003 Annual Meeting CD-ROM Paper revised from original submittal.