Research On Urban Traffic Data Fusion And Traffic State Processing Methods

The traffic congestion has become one of the main problems in the city all over the world. With the rapid growth of our national economy and the unprecedented speed up of urbanization, the number of automobile increases rapidly year by year, which result in traffic-jams in many cities. Even if many cities have increased their input in construction of traffic facilities, it still cannot resolve the contradiction between the increasing need of transportation and the low utilization of traffic facilities. Intelligent transportation system (ITS) has been seemed as the best acknowledged method to solve problems in traffic field at present. Traffic data fusion and traffic state processing methods are the base of ITS and provide an index for urban traffic.This paper firstly introduces the concept of floating car, map-matching and some other technologies about traffic state processing, including floating car data (FCD) collection and preprocessing methods, determination of the number of floating car, support vector machine prediction, traffic state evaluation system and GIS digital map development. Based on the analysis of map-matching algorithms, the road network of Jinan was built, and we design a large-scale increment precise map-matching algorithm(LSIPMM) to increase the matching accuracy and a traffic state processing algorithm based on average velocity with support vector machine to predict and evaluate the level of the traffic state.After the traffic state has been obtained, J2EE technology is used to display the traffic state in web pages. The FCD was used to test and simulate, and the pages of the result were showed in the end.