Networked Spreading Characteristics Of Traffic Congestion Caused By Typical Traffic Incidents

Traffic congestion problem brought by the rapid urbanization has become one of the main social problems in the metropolitan cities of the world. Especially in recent years, vehicle ownership has also increased dramatically, and the existing urban road network is difficult to meet the traffic demand, so the urban road congestion is becoming a more and more apparent problem. A major source of this problem on the urban expressway is the non-recurring congestion caused by various unexpected incidents, which is extremely easy to generate the bottleneck of traffic capacity. Existing research of the road congestion spreading mainly focused on the prediction of the changing process of queue length. However, the traffic flow characteristic on the urban road networks is very complex due to the physical link topology, and the spatial and networked spreading characteristics of the road congestions, which necessitate the study on networked spreading characteristics of road congestion based on existed traffic flow data.In this context, this thesis investigated the networked spreading regularities of road congestion caused by typical traffic incidents. On the basis of the quantitative analysis of congestion spreading velocity on basic segments of urban roads, a visualization computer program has been developed by using Geographic Information Systems (GIS), in which the network topology relationship among the roads bas been embedded. First, based on traffic shock wave theory, this thesis considered congestion spreading as a rally wave transmission process. Then the congestion spreading velocity under the typical traffic incidents was calculated by taking advantage of Van Aerde traffic flow model. The second, traffic flow characteristics were analyzed from three aspects including different times, different sites and different traffic incidents. Moreover, the road congestion spreading velocity is analyzed and simulated. And based on the analysis, the congestion spreading velocity model was established which was suitable for describing the spreading characteristics of congestion caused by typical traffic incidents in Beijing. In addition, the congestion spreading velocity database was established for various road types and time periods. The third, on the basis of the research of physical connectivity and topological relations of road network, the effect of intersections on traffic flow characteristics has been analyzed, and then networked spreading characteristics of road congestion for typical traffic incidents were concluded. Finally, the visualizing display system of networked congestion spreading was established, in which the system database, system framework and functional structure were designed based on a GIS platform and C#development environment. Case study indicated that the system was practical and effective, and was able to provide intuitive road congestion spreading state for traffic manager, and provide decision support for developing the corresponding strategies of traffic management.