Continuous Congestion-based Comprehensive Measurement Approachfor Urban Traffic Congestion

With the near completion of the formation of urban road networks, traffic congestion occurs more frequently because of the increasing conflict between the large travel demand brought by the rapid growth of automobile population and the limited road resources. The public begins to pay much attention to traffic congestion problems while enjoying the advantages of rapidness and convenience of automobiles. Traffic congestion problem not only becomes an important element prohibiting the harmonious development of the city, but leads to serious economic losses and causes a series of environmental problems such as noise pollution and air pollution as well. This phenomenon is especially apparent in large and mega-cities. In order to manage traffic congestion more practically, reasonable systematic evaluation of the impact brought by the traffic congestion becomes the important premise and basis of congestion management. In this context, this thesis is intended to conduct the following research:First, a comprehensive traffic congestion estimation model based on "continuous congestion" is proposed, through an analysis of traditional traffic congestion evaluation system in a combination with the four-dimensional characteristics and formation-dissipation characteristics of traffic congestion. The most important feature of this model is its use of the formation-dissipation process as the evaluation target, which sufficiently reflects the concept of centralization in the congestion evaluation, completely changing the averaging evaluation approach that has used the road classification as the basis. The proposed model centralizes the congestion evaluation, which breaks the restriction of road classes and regions, thus expanding the comparable range of congestions. Further, it proposes a comprehensive congestion index system by combining the four-dimensional characteristics of traffic congestion, which introduces the conception of virtual vehicle flow to evaluate the impact of continuous congestions from the perspectives of either travelers or managers based on the different levels of demands.Second, the thesis studies two key techniques. The first is the evaluation of the traffic congestion intensity. The thesis adopts Fuzzy C-Means algorithm to divide the speeds on different road classes according to the perception about the congestion intensity using the traffic congestion intensity survey. Second, the thesis proposes and designs identification method of a continuous congestion by combining with the technology of geographic information system, and realizes the identification and comprehensive evaluation of a continuous congestion in different temporal- and spatial- levels through the secondary development of ArcGIS software.Finally, a case study is conducted to implement the proposed method and theory by using the floating car data, microwave sensor data and geographic information data, followed by an analysis of the identification and evaluation results of a continuous congestion. The study on the west second ring road network of Beijing shows the feasibility of the proposed evaluation model, demonstrating the practical significance of the comprehensive evaluation of "continuous congestion". Suggestions for improvement are also provided to relevant parts during the process of practice.