Study on the Traffic Flow Guidance Model of Expressway Based on Data Fusion

Aiming at the requirements of the complicated structure of the express highway circling centre city, and the high dependent degree of the traffic flow guidance, a model of the travel time prediction is proposed which based on the data fusion of the dynamic real-time collection data and the historical traffic data. According to the real-time collection data, one prediction value is calculated through the field collection traffic parameters which were fused beforehand, another is prediction from the historical data by using the improved gray model (GM). The final prediction value is fused by the former two types of data. A basic process for this model is proposed which can be applied to predict other traffic guidance information besides the travel time prediction. The effectiveness of this process and travel time prediction were tested using detected data collected in Chengdu city.