Simulation Research On Urban Traffic Flow And Intelligent Control Of Crossroad Signal

In this paper, we do some research in simulation of urban traffic flow and intelligent control in crossroad. First we build traffic-flow model based on Cellular Automata, and use fuzzy control to build lane-changing model, because fuzzy control adapts to dealing with subjective judgement decided primarily by man. Combining fuzzy control and neural network makes the best of their advantage, we build intelligent signal control system in crossroad based on fuzzy neural network. This control system has some characteristics. Firstly, we consider not only traffic status of one crossroad, but also traffic status of backward crossroad. Secondly, we adjust prolonged time of current green phase and choose next green phase. Thirdly, we use fuzzy neural network to adjust membership function in fuzzy control, which makes membership function accords with fact. Fourthly, during simulating process, we use systemic data to train fuzzy neural network many a time, which optimizes network.We design simulation system with Visual C++ for traffic-flow model and lane-changing model and intelligent control of fuzzy neural network. The simulation system has the following performance, â‘ visualization: Users can see the moving status of cars on the road and in the crossroad, â‘¡applicability: The simulation system can get the data of car flow from the system itself, and also can get data from real-life. It can adjust the parameter value automatically based on the analysis of intelligent control module, â‘¢controllability: Users can change the parameter value of lane and each crossroad, such as the length of lane, the cycle time of crossroad and the number of phase of crossroad. These performance make users watch and analyze the status of traffic easily, and can also help users work out the effective decision, ensure the traffic flow fluently.Through simulating, we can see that vehicle's moving status exactly reflects traffic status in fact. At the same time, the simulation system validates the intelligence control method too. Experimental data shows that the intelligent control method iseffect in crossroad traffic-flow control.