Abstract:

This paper proposes a new dynamic traffic congestion prediction model. In this model, the probe vehicle data are used to estimate the real-time traffic flow situtation in the road network, and the static topology structure of the road network is combined with the multiple fuzzy inferences to predict the possibility, degree and occurrence of traffic congestion. Practical data show that the proposed model is of high prediction accuracy.

Key words: traffic congestion prediction, short-term traffic forecast, probe vehicle technology, intelligent transportation system