Model of Road Traffic Accidents Prediction Based on ARIMA-FNN Optimal Weighted Combination

A good adaptability road traffic accident prediction model is not only important for traffic management and control but also for traffic plan.According to the space-time nature and multi-factors of traffic accident,this paper established a combination model with ARIMA and FNN,which is based on the way of synergic curve weighing to forecast the traffic accidents in China.In this model,the ARIMA can linearity fitting time sequence and FNN has the great ability of nonlinearity mapping.The prediction results show that the combination model is an effective traffic accident forecast model,which can increase the precision of traffic accidents prediction.