The Design Of Intelligent Control System Of City Traffic Signal Lights Based On Neural Network

Urban traffic signal control is an important component of Intelligent TransportationSystem. Today’s traffic signal light control strategy is basically fixed which is difficult todynamic control the changing traffic, causing tremendous traffic congestion and enormousenergy consumption. In some complex traffic situations, using the traditional methods tocontrol the traffic signal is not very good. Because of the developing economy, road andcurrent transportation equipments can’t satisfy the needs of the existing automotive growth,how to use simple and economical way to improve traffic efficiency is an urgent need toaddress the problem.Neural network is a system, it has the characteristics of highly parallel, non-linear anda high redundancy. The neural network has the self-learning and self-organizing abilities.This is a unique advantage for those who can not establish accurate mathematical model ofthe system.Based on the complexity, changing, random of urban traffic intersections, trafficsignal control were analyzed by the method of application of neural network control. Thispaper describ a design approach based on traffic flow changes dynamically to adjust trafficlight time. The green time of the lane is distributed by detecting the on-site traffic, andaccording to the neural network algorithm in order to achieve dynamic traffic regulation. Itwill improve the capacity of the intersection effectively. This intelligent control of trafficlights can make the traffic be controlled effectively, it has a significant effect to diverttraffic, improve road capacity and reduce traffic accidents.