The Research And Application Of Traffic Flow Detection Method Based On Fusion Of Un-congeneric Information

The real-time traffic information is the foundation of ITS (IntelligentTransportation System), dynamic traffic flow is the key of ITS researching. With thedevelopment of ITS, many kinds of techniques are created for collecting the trafficinformation.Those techniques have their own advantages and disadvantages, sovarious detectors are used on the road, then researchers have many different trafficinformation and it is hard to get the real traffic condition. How to deal with the vasttraffic data and get the real-time, total, credible traffic flow is the ITS researchemphasis.After studying the researches at home and abroad, firstly, the RFID traffic datasimulated by traffic micro-simulation software VISSIM, and be proved theeffectiveness of RFID data in traffic flow detection and estimate with the kalmanfiltering algorithm. Secondly, because the floating car detection and RFID detectionare all vehicle detection, so they can be managed together through the kalmanfiltering and Dempster-Shafer evidence theory, then obtain more comprehensivetraffic flow. Finally, according to study the secondary development of ArcView, thetraffic flow detection and analysis module was designed and implemented bySQL2005、VS2008and MATLAB. The module can show a city’s traffic flow visually,for information fusion methods are inset in the module, it can display the differentflow in different algorithms.This paper lucubrates the traffic flow theory, traffic flow model, the principleand application of RFID technological; studying the data fusion algorithm ofheterogeneous multi-sensor system, geographic information system software andmicroscopic traffic simulation software; according to research the arithmetic andrealize the traffic flow detection module, the result shows that the method iseffectiveness and the design is feasible, this paper’s research afford a importanttheory support for RFID in traffic flow detection.