基于agent的智能交通控制的研究-控制科学与工程·控制理论与控制工程专业论文.docxVIP

AbstractABSTRACT Abstract ABSTRACT With the development of the economy,congestion in urban traffic has been a problem commonly concemed around the world．Traffic congestion will cause the increase in traffic delay,traffic incidents，fuel consuming and so on，and thus bring great social problem and tremendous economic loss．In this dissertation，in allusion to the status quo of the urban traffic and based on agent technology which is new paradigm of computer and artificial intelligence，the intelligent traffic control problem is discussed and studied，which is expected to be an effective method to solve the traffic problem． Firstly,the developing status of the agent technology and multi—agent technology is introduced，and the coordination technology of multi—agent is summarized． Secondlythe basic methods and the basic structures of traffic control are introduced， based on which，an agent based urban traffic signal control system structure is presented．This system is a three—layer structure，and includes four kinds of agents： intersection agents，road segment agents，area control agents and control center agents． In this system，the traffic control strategy is made by the intersection layer,related agents communicate with each other,and thus the reliability,the real—time flexibility and the expandability are improved．Thirdly,the model structures of each kind of agents are designed in detail，and the specialties，functions and working steps of each kind of agents are described and analyzed；with introducing intersection queue matrix， road factor matrix，traffic load matrix and green signal adjusting matrix，multi-agent coordination to intersection congestion and road special incidents is modeled，and coordination algorithm is put forward．Fourthly,after analyzing the traffic flow of urban intersection，a four-phase traffic signal control model，taking the average traffic deley as the performance index，is modeled，and an improved genetic algorithm is applied to this model．The simulation res