A framework of real-time traffic information system

In this paper, a framework of real-time macroscopic traffic simulation is proposed. Real-time traffic simulation using macroscopic traffic flow model on freeway have been studied and implemented in the past. We introduced a framework of implementing real-time traffic simulation on urban network. To implement the simulation on an urban network, the estimation of origin-destination (O-D) matrix must be considered. Thus the framework consists of 1) Path generation, 2) Dynamic Origin-Destination Estimation and 3) Link Dynamics. Path generation is based on the Wardrop's user equilibrium concept and modified convex combination algorithm. State space model is used in dynamic O-D estimation, while the LWR model is used to describe the macroscopic traffic flow in Link Dynamics. This study implements the simulation with real-time traffic input and predict traffic condition in real-time. An empirical example is conducted on a real network consists of 91 nodes and 244 links.