Prediction of Urban Traffic Abnormity Based on Causal Network

Traffic abnormity seriously impacts the normal operation of urban transport system. The detection of traffic abnormity is a passive mode and of limited effectiveness. A prediction method of traffic abnormity is proposed, which can be a proactive response mode. Three parameters are brought in to describe the traffic abnormity. ARIMA is applied to predict these three parameters first. Taxi GPS data is used to train the neural network and predict the traffic abnormity on the dimension of time. Because of the background probability, the results of basic predicting model need to be modified more precisely. Causal network is employed to revise the predicting results on the spatial dimension. Results of Pearson test shows that the modified predicting results of traffic abnormity could be acceptable with the confidence level 80%.