Congestion Forecast Model from Integrated GPS/GIS Data Based on Fuzzy Logic and Neural Network

In dynamic traffic management (DTM) congestion situation forecasting is most valuable information. So far, most the congestion forecasting models are developed on point based traffic data. Yet, in developing countries like China, stationary detectors are very limited and can't support the DTM. Thus the paper presented an adaptive on-line congestion forecasting model from Integrated GPS/GIS data based on fuzzy logic and neural network. This model produces estimates for congestion based on the history and on-line GPS/GIS information. The performance of the model shows good results when compared with real data in Guangzhou city. This study would contribute as a basis for applications of the GPS/GIS based DTM.