The Study Of City Traffic Signal Control System Based On Agent

The traffic problem has been a focus. It has been great influencing the operation of society and economy. City Traffic Signal Control System is the way of improving city traffic operational efficiency and flag of city modernization. Excellent CTSCS can improve city traffic problem.First, this thesis analysis development history and development tendency of traffic signal control system and introduces main parameters and evaluating indicators of traffic signal control. It studies specially the prediction method and model vehicle-flow.Second, this thesis analysis current traffic control system structure and brings a new city traffic signal control system structure based on Agent. The system hands the authority of deciding on control strategy to intersection. The intersection controls traffic signal according to itself knowledge, through cooperation. Intersection improves capability of intellect strategy. So control system can adapt flexibly to complex features of traffic system. It supplies a probability for real-time intellect control.Goal of this thesis forwards a new real-time self adapting city traffic signal control system based on Agent Technology. Thesis analysis Agent traffic signal control model based on general knowledge. It studies learning mechanics based on Genetic Algorithms and cooperation mechanics based on Game Theory. It emulates mainly the traffic model of high vehicle-flow. The result of emulation shows that total control effects are better traditional controls'.