Combined Forecast Model for Medium-term Traffic Flow Based on Polynomial and Fourier Series

In order to improve the accuracy of medium-term traffic flow prediction, a combined forecast model for traffic volume was established based on polynomials and Fourier series. In the model, the prior information of traffic flow in flat hour and peak hour was used through polynomial trend regression analysis,and Fourier series was adopted to repair the impact of random factors on traffic flow. Finally the stability and effectiveness of the method were tested by a case study using the traffic flow data of road Wapenyao in Harbin.The results show that the method is of high prediction accuracy, and is generally superior to all kinds of exponential smoothing models such as ARIMA,BP neural network, RBF neural networks and other typical short-term traffic flow prediction methods.