Study on Short-term Traffic Volume Prediction Model
Song Zi-fan · Scientific Decision-Making · 2014
To solve the road congestion and make proper road safety planning,relevant departments should strengthen the road traffic flow's real-time monitoring and forecasting. Only in that way,can we detect current abnormal traffic congestion and improve travel efficiency of the people. The characteristics of short-term road traffic flow are uncertainty and nonlinear. For these characteristics,first of all,we use the gray system theory and time series forecasting model of ARIMA respectively to forecast the traffic flow. On this basis,we propose a new foresting model with combination of both methods. By analysis of the comparative examples,we conclude that the prediction accuracy of the combination forecasting model is higher than gray prediction model and time series analysis model. The combination model can be used as an effective method for shortterm trafficvolume forecasts.