The Intelligent Traffic Light Control System Based On Information Fusion Research

At present, the urbanization of our country develops very fast, The number of vehicles also rises rapidly. The contradiction between urban traffic and road is more and more intense. And the controlling methods of traffic lights lead the low utilization rate of the road. At a result, more and more traffic accidents happen on the road. So in order to prevent traffic congestion, the exploitation of an intelligent traffic light very deserves research. At the same time our country is increasing the investment in intelligent transportation systems. This system provides a broad prospect about the application. At the same time it provides a support about the hardware. For all these reasons, this paper puts forward a kind of intelligent traffic light controlling system. It based on information fusion technology. It helps traffic policemen organize and manage traffic vehicles reasonably. Then we can reduce the traffic accidents.This paper proposes a forecasting method based on cusp catastrophe theory to predict the traffic jams. At first, we use the wavelet neural network to predict the traffic information based on the collected information of vehicle speed and vehicle number. Then using traffic prediction method based on cusp catastrophe theory for road traffic density and speed of road traffic information for information fusion, given road traffic congestion critical vehicle density and critical vehicle speed, Finally by predicting vehicles information and it is concluded that the critical information for the optimal vehicle scheduling strategy. In this paper, based on information fusion of intelligent traffic light control system’s overall architecture, functional architecture and technical architecture design and signal control simulation of scheduling module.Experiments have been carried out to test in this paper, the actual path of the ve hicle, and probes into the actual road vehicles in the intelligent traffic scheduling. The experimental results show that through the wavelet neural network for neural network training, the first three days to forecast the traffic flow situation of the fourth day, th e traffic flow prediction results and the actual observation result is found that more th an 84% accurate. Through this intelligent light control system scheduling after the inter section traffic flow extremum and the original 300 fell to 185 now, ease road traffic c ongestion by an average of 27.89%, the intelligent light control system play a more efficient scheduling effect, can be applied to the actual traffic control, the system develo pment provides the foundation for the next step.