Study of traffic flow detection method based on Nearness Degree Data Fusion Technology

In the traffic information collection, Multi-sensor can be achieved in time and space information complementary. How to extract and process the discrete data effectively, become an important issue on multi-sensor comprehensive information collection. Based on the principle that same traffic conditions lead to similar traffic status, we present a new multi-sensor detection method for traffic flow by applying Nearness Degree Data Fusion Technology. According to the actual traffic measurement background, in the article we also put forward another improved idea that time factor which has major impact on should be considered into the actual engineering application of the traffic flow detection. Firstly, with reference to noise ratio method of multi-sensor data fusion, the traffic sensor measurements collection is treated as a fuzzy set, it measures the mutual integrated nearness degree of sensors at different times by their maximum and minimum close levels. Then the multi-sensor data fusion methods and algorithms for traffic flow detection based on the defined consistent reliability measurement were designed. At last, an application example experimental was performed by Matlab programming and it proved that the proposed method is simple and effective. The method has good accuracy and stability.