Road Congestion Analysis in the Agglomeration of Sfax Using a Bayesian Model

This study provides a road traffic portrait in urban areas to compare the congestion level of certain sections. In view of a better exploitation, we proposed a Bayesian network (BN) analysis approach to modeling the probabilistic dependency structure of congestion causes on a particular road segment and analyzing the probability of traffic congestion. In this case, two steps are also necessary, the macroscopic traffic flow modeling and the traffic simulation for which empirical measurements can be developed and tested. The BN method is used to analyze the uncertainty and probability of traffic congestion, and is proved to be fully capable of representing the stochastic nature of road network situation. This approach is used to represent road traffic knowledge in order to build scenarios based on a practical case adapted in the city of Sfax.