The Traffic Flow Forecasting And Signal Contral Of Urban Single Intersection

As the converging point of the city road, single interaction play a very important part inthe whole city road network, especially in the main road of the city. The control effect of thesingle interaction strongly affect the traffic efficiency of the city road network. So it’snecessary to take a deep research on the single interaction.This article mainly contains two parts, first of all, according to the traffic flow data andthe netural network theory, design the modle of BP netural network, then train and simulationthe network. According to the simulation result and the function in MATLAB, devide the trainsimple into three groups, and then creat a improved BP netural network.According to the fuzzy control theory, and with the help of fuzzy logic box, creat atwo-class fuzzy control system. Asure the queue length calculation formula in different lightof the single interaction.Improve the input of the fuzzy controller, an important err that ismade of the maximum value of the queus length of the red phase and the value of the queuelength of the green phase. Compaired with the classic fuzzy controller, and make sure thetwo-class fuzzy controller is better than the calssic one.