A SEMI-DYNAMIC ASSIGNMENT MODEL CONSIDERING SPACE-TIME MOVEMENT OF TRAFFIC CONGESTION

Abstract

Traffic condition in most cities varies significantly within a day, in which static traffic assignment model may not be able to sufficiently represent time-varying congestion phenomena in transport network analysis. Different semi-dynamic traffic assignment models have been proposed in the past to tackle this pitfall. However, most of the models still face the serious problem of space-time propagation of the traffic flow and congestion. In this study, we propose the model which can model space-time propagation of the traffic congestion. In the proposed model, the traffic volume on a certain link will be carried over to the next time period representing the propagation of the congestion from one period to another. The paper then analyzes the existence and uniqueness of the model's solution. The model is then tested and validated with an example network in which the model result is compared against those from the previously proposed models.