A Method Of Forecasting The Traffic Accidents Based On ANN

With the increasing of the highway mileage and vehicle quantity, the traffic accidents which is a menace to our life and property safety become more and more, and this trend is going up, it has become a severity societal problem which we should pay full attention to. The road accidents forecasting is an important research content of the traffic safety. The purpose of road accidents forecasting is to analyze the tendency of road accidents under existing road traffic conditions, estimate the feasibility and practical effectiveness of traffic safety measures, control the factors affecting road accidents, and reduce the traffic accidents.In this paper, the significance and complexity of the road accidents forecasting are studied; then some familiar methods to this question and their insufficiencies are discussed. According to the characteristic that the traffic system is dynamic and time-variance and the advantage that ANN has in solving complicated and nonlinear problems, a method of forecasting the road accidents with BP ANN is proposed, and a forecasting model based on BP ANN is made; finally traffic accidents, death tolls, casualties and direct economic losses and other parameters related to traffic accidences are forecasted. The results prove that this method is feasible, having high-predicted precision and constringency speed, it provides some valuable references for the traffic management and controlling.