View article

During the past two decades, considerable research eorts have been devoted to the development of dynamic trac models mainly motivated by ITS (Intelligent Transportation Systems) technologies. These models describe the evolution of the congestion of urban and inter-urban networks as a dynamic process instead of considering-as in static models-that this phenomenon is stationary. This dissertation presents an innovative methodology that uses the simulation approach. In the rst part, we build a fully dynamic model with two key features:(1) a mesoscopic event-based trac model and (2) a departure time choice model based on the work of W. Vickrey. Its advantages are the ability to compute stationary time-dependent states and to take into account within-day time as well as day-to-day adjustments. Computer implementation and performance issues are addressed to allow to simulate large-scale networks within …