Modeling and Forecasting of Short-term Traffic Flow Based on Grey Model with Oscillation Factor

Short-term traffic flow prediction should have the features of real-time, accuracy and reliability. The prediction result directly affects the traffic control and induction. Therefore, it is the key technology for the advanced traffic management information system. Considering the uncertainty and periodic oscillation of traffic flow data, a new prediction model of short-term traffic flow has presented based on grey model with oscillation factors. Periodic and model parameters have solved by using discrete Fourier transform method and least square method respectively. The new model significantly improved the prediction accuracy and requirement of the real-time prediction, comparing with the predicted traffic flow value of the GM(1,1) model.