Studies On Congestion Propagation And Control Strategy Of Urban Traffic Based On Complex Networks

Urban traffic network is the basis of urban prosperity. It's important for residents' daily life, logistics transportation and economic development that urban traffic network keeps working normally and efficiently. However, with the urban development, the supply from urban traffic network gradually cannot match the growing demand. Urban traffic congestion has made a serious influence on people's living standard, restricted urban development and triggered a series of social and environmental problems. To prevent and mitigate traffic congestion, the making of traffic plans, traffic management and traffic control strategies should be guided by the mechanism of traffic congestion formation and the principle of its propagation and dissipation in the perspective from the whole network.Urban traffic system is an opening complex giant system. Its running is influenced by many factors. Most of previous studies simulated the vehicle transmission on it by using the traffic flow theory based on the simple grid network. But they cannot fully reflect the complex characteristics of the urban traffic network. Recently, complex networks theory has developed quite a lot. It focuses on analyzing the relationship between the network structure and its dynamic behaviors. It really offers a new perspective for studying urban traffic congestion problem. However, most of the applications which use the complex network theory studying urban traffic congestion, adopt the static user equilibrium model. As cannot reflect the dynamic evolution of the congestion stretch among different traffic state, it is not conducive to find the law of traffic congestion propagation.So in this paper, considering the actual characteristics of real traffic network, the urban traffic network model is established based on the complex network theory. By adjusting the classic cell transmission model, the evolution process of traffic flow on different supply-demand structure is simulated, the space-time characteristics and transmission mechanism of network congestion are analyzed. Then an effective congestion control strategy is proposed. The major contributions are as follows:(1) According to the actual characteristics of the real urban traffic, the urban traffic network is divided into the upper demand network and the bottom road network. Comparing the evolution characteristics with typical complex networks, the upper demand network is abstracted to a scale-free network created by the fitness model, and the bottom road network is abstracted as regular network, small world network, random network and grid network respectively. It will be more realistic in reflecting the characteristics of real urban traffic network.(2) For the need of the study, the cell transmission model is adjusted to combine the upper demand network with the bottom road network by setting the vehicles' origins and destinations. Moreover, the routing choice of every vehicle is made dynamically according to the dynamic traffic environment. And both the hysteresis effect of road traffic and road channelization at intersections are considered in the model. Those make the loading and the movement of vehicles in the model closer to the actual situation.(3) By using the vehicles creating rate of every node to show the traffic demand intensity, the scale of traffic congestion and the delay caused by traffic congestion are studied on different network models with different demand intensity combined with the topology characteristics of the upper demand network and the bottom road network. And then, the distribution characteristics of traffic volume on different road networks, the evolution process of the network congestion and the development and space-time characteristics of traffic bottleneck are analyzed under different traffic demand intensity.(4) After analyzing the relationship between network topology structure and traffic flow evolution, the cumulative betweenness strength is defined. It is found out that the cumulative betweenness strength could forecast the trend of edges' traffic volume in some degree. So adding it into the calculation of traffic impedance, the betweenness protection strategy for traffic congestion control is proposed. The strategy would diminution the influence of traffic congestion by homogenizing the traffic flow on road network and avoiding the excessive attraction of roads. And then, the effectiveness of the betweenness protection strategy is verified after computer simulations.