Structure Characteristic Analysis Method Of Road Traffic Network And Optimization Of Typical Intersections And Road Sections

Road traffic network is the basic skeleton of the city. Its role is the public travel and cargo transport which ensure the normal operation of the various fundamental function of city. It can promote the economy and social progressive. But the traffic congestion in cities have become more and more serious. It restricts the development of urban economic and social civilization. At the same time it seriously affects the quality of life of urban residents. And the traffic congestion is the most insidious effect of the traffic safety. Modeling and structure characteristic analysis of urban road network can provide a certain amount of supporting data for the urban road designers and the city traffic managers. Considering the fixed location and regular time of the urban traffic congestion, the identification of the key intersections and road sections and the time series analysis of road traffic network states can help the traffic managers generate the targeted strategy to ease traffic congestion. The optimization analysis of the typical intersections and road sections can improve the traffic efficiency and ease the traffic congestionBased on the background of easing traffic jams, this dissertation studied the modeling approaches, the structural characteristic analysis method, the identification method of key components, the time series analysis of traffic network states, the optimization analysis of the typical intersections and road sections. The part roads network in Beijing was used as an example to prove the effectiveness of the methods. The main research achievements are embodied in the following five aspects:(1) A pervasive model of road traffic network was built. The model can form three network models which include the network model without any attributes, the network model with static function attributes and the network model with the dynamic function attributes. The indexes of structural characteristic analysis for the above three network models were provided separately. The part roads network in Beijing was took as an example to validate these indexes.(2) Considering the fixed location of the urban traffic congestion, this dissertation proposed the identification method of key intersections by considering node connectivity degree, adjacent nodes connectivity degree change caused by node failure, node average shortest path distance and network average shortest path distance change caused by node failure based on the network model without any attributes. The identification method of key intersections by considering node length degree, node capacity degree, node efficiency degree and node betweenness based on the network model with static function attributes was also provided. The identification method of key road sections by considering edge real-time flow, edge real-time velocity, edge real-time occupancy and edge real-time state based on the network model with dynamic function attributes was put forward. The identification method of key road sections based on adjacency relations and time distribution of traffic congestion was presented. The part roads network in Beijing was took as an example to analyze the above four identification methods of key components.(3) Considering the regular time of the urban traffic congestion, the time series analysis of traffic network states was provided with considering the flow, speed and occupancy. The method first used the betweenness to calculate the importance of components, secondly obtained the time period of morning and evening peak traffic state by using the state division of components based the optimum partition method, then got the time distributions of morning and evening peak traffic state based on weighting method of time period, finally gained the trapezoidal fuzzy time period of five traffic states through probability-possibility transformations and possibility theory. The part roads network in Beijing was took as an example to describe the process of the time series analysis. The example results was discussed and analyzed.(4) Aimed at improving the traffic capacity of the intersection based on the optimal control of the traffic signal lamps, the traffic signal real-time control1 model was proposed based on fuzzy hybrid Petri net. This model not only can compute the best green light duration of each phase through queue length and intersection traffic flow but also can determine the reasonable and alterable phase sequence through red light duration and queue length. The simulation analysis of a simple traffic signal with four phases was provided.(5) Aimed at improving the traffic capacity of the expressway based on the optimal design of the bus lane, this dissertation proposed a new bus lane with limited physical length. Based on the cellular automata traffic flow model and the lane-changing behavior of the vehicle including the optional lane-changing and the mandatory lane-changing, a three-lane traffic model with an isolated no-bay bus stop was provided. The ordinary three-lane traffic without a bus lane and the cases of traffic with a dedicated bus lane or the proposed bus lane are simulated. The comparisons and analysis are made among them.