Research On Theory And Method For Urban Traffic Flow Guidance System Coordinated With Traffic Control Based On Multi-Agent

Presently urban traffic problem is a globally puzzle, especially in China. In which urbanization becomes more and more with great economic development. Serious traffic congestions and frequent traffic accidents always exist. The extension and construction of roads cannot solve the problem. As a means of improving urban traffic, Intelligent Transportation Systems (ITS) obtain more and more recognition especially in China. The traffic management system, as an important constituent of ITS, greatly benefit and improve traffic management. Urban Traffic Control System (UTCS) and Urban Traffic Flow Guidance System (UTFGS) are two important subsystems of Intelligent Transportation Systems (ITS). The cooperation of the two subsystems is the key method of resolving the traffic problem. For the Characteristics of modern traffic are complex, levity and distribution, the Multi-Agent Technology is adopted in the study of the urban traffic flow guidance system coordinated with traffic control , and the Intelligent Decision Support Technology is used in the global solution optimizing, in order to reduce the traffic jam, decrease the exhaust gas, improve the efficiency of the traffic management.This research consists of six chapters. The main works and contributions are described as follows:The main contents as follows.1. Introduction. Chapter one mainly summarizes the research background, the aim, the meaning and the content of this thesis. Through the analyzing of the present study and the approach of the Traffic Control and the Traffic Guidance cooperation, find out the exiting problem, and summarized the cooperation model and the arithmetic.2. The study on the framework construction of the cooperation of the UTCS and the UTGS. Combining the complex characteristic of the traffic and the Multi-Agent, analyzing the feasibility of the Multi-Agent technology applied in the cooperation of the two systems. The distributed hierarchal architecture of the cooperative implementing system is presented. The Intelligent Decisions Support System theory is also adopted for affording the basis for the globe solution established. There are the Section Agent, Control Agent, Cooperation Agent, Area Agent and the Centre Agent in this system, All the Agents Share the common information, and the upper Agent is charge of the lower Agent.3. The information processing of the Traffic Control and the Traffic Guidance cooperation.In traffic management, timely access to accurate, comprehensive traffic information directly related to the correctness and rationality made by traffic management strategy. Travel information which can token traffic flow parameters and react traffic state is the basis of achieving traffic control and traffic management strategy. Only based on real-time, high-precision basic traffic parameters information and network traffic flow state information,can we formulate the corresponding traffic control and induced strategy .then manage traffic flow to the best running reasonablely and effectively.collection and processing of transportation infrastructure information is the guarantee of achieving the Traffic Control and the Traffic Guidance cooperation.after traffic information is collected and processed, it provides traffic control and traffic-induced collaborative applications.This chapter focuses on the information processing technology of the Traffic Control and the Traffic Guidance cooperation,and proposes a fusion algorithm used in the space traffic speeds.The algorithm can analyze automatically and treat integratedly the detected traffic parameters (traffic flow, average speed, instantaneous speed, distance at the front, the model classification, lane share, etc.) in a certain criteria, then complete the necessary decision-making and assessment.it can coordinate multi-source data reasonsablely, integrate useful information fully,increase the correct decision-making capacity in the volatile environment.4. Research on traffic control and optimization of induced synergies based on the multi-agent.presently, traffic control and induced synergies has been a national focus on transport workers.it refers to under the macroeconomic regulation and control in the centre of traffic, based on different traffic flow, make full use of the advantages of complementarity between the junction, balance traffic flow of each intersection, So as to enhance the capacity of the road. This chapter mainly research on how to use the multi-agent means to achieve control and induced optimize synergies.apply strengthening the study to realize study characteristicof multi-agent.combining genetic algorithm, establish the model of traffic control and induced optimize synergies and solve algorithms.with VISSIM software, simulate using analog network,verify the effectiveness of the algorithm.5. Research on overall optimization decision. This chapter analyzed the foundation of the overall optimization strategy. Because the coordination of traffic control and traffic navigation confront to larger regional context, based on multi-agent's self-government, before the overall optimization, the whole region is divided into several sub-region, and the complex tasks of the coordination of traffic control and traffic navigation is divided into several sub-tasks and the overall optimization strategy is easily achieved. This chapter introduced knowledge-based model, which involves the acquisition of knowledge, repository and reasoning mechanism, hold the experts' knowledge and experience in a particular field, and can use the knowledge like the experts, and make intelligent decisions by reasoning in the field, and simulate the process of decision-making to solve the complex problems which can be solved by the experts. At the same time, combine the characteristics of multi-agent study, apply Q-learning to establish the intersection model and the model of regional coordination in traffic control. According to the decision-making strategy of winning or losing, Game Theory will be applied to strategic planning of decision-making level the overall optimize and give an evaluation index of the model and system.6. Concluding remarks of the research. The main research achievements and further study problems are pointed out.