Traffic Flow Analysis Of Urban Road Network

The urbanization process that swept the globe has expedited the pace of city life. The trend to travel with vehicles is prevalent throughout the world, especially for modern metropolis. As a consequence, there is a growing number of contradictions within the systems of urban traffic. In recent years, the rapid development of ITS(Intelligent Transportation Systems) has become a key method to address this problem, where a reasonable and accurate analysis of traffic flows in urban road networks is a vital part.Inductance loops and floating cars equipped with GPS are useful and economical sources to detect traffic flows, thus are widely employed in researches and applications of ITS. Using these two sources of traffic data, we first implemented the macro traffic-wave theory and the algorithm of surface fit to estimate the traffic state of urban road networks in Shanghai.Then we put more emphasis on GPS data, and proposed a vehicle tracking method to estimate traffic states. The method dynamically links real-time GPS locations of floating cars to track their paths in the urban road network. An by calculating the average speed of these tracks, we managed to estimate the traffic state for the whole road network. The merit of this new method basically lies in that it is not affected by the lack of information upon crossroad waitings.Detailedly, we introduced Pareto multi-objective optimization and fuzzy theory to comprehensively judge the most likely path for each tracking problem. During this process the driving distance and the number of passing crossroads were set as two optimal objectives. After that, we distributed the average speed of each path proportionally to the road segments it covers, and integrated all the speed contributions a road segment collects for a final estimation of its traffic state.In addition, we discussed about a combination of federal filters and the evidential theory to realize information fusion for traffic data on a characteristic level.The whole process of data acquisition, analysis, fusion and display are all integrated in the traffic information processing platform we developed based on GIS. It regards a single directional road segment as the basic analyzing unit, and divides the degree of traffic congestion into five levels for display purpose. Further applications such as traveler guidance and traffic information broadcast are hopefully expected.With analyses upon real traffic data, we concluded that the fusion of GPS and SCATS data gives better estimation results for urban traffic states in that it enables mutual supplement between different data sources and evaluates reasonable degrees of credibility according to their behaviors.