Exploring the use of DTA for origin-destination matrix estimation

Traffic simulation is more and more widely used and a helpful decision tool for planer and managers. Several tools exist and could be grouped in two main families: Macrosimulators with aggregate vision of the traffic in the network and microsimulators which provide individual data. These two tools got different approaches of modeling traffic and could be use for different applications, but in some case, links or interactivities are observed depending on the inputs needed. Unfortunately and due to the differences between models, these links are not lacking of compatible problems. This paper presents a study with an innovative dynamic approach of the OD matrix determination process comparing to the usual static one. Evaluation and tests of the proposed method are presented to quantify the improvement. Using dynamic traffic assignment, developing a historical data base representing the traffic network behavior during time and finding a new matrix adjustment fitting with the data used, are the main innovative contributions of this work.