Study on short-term forecasting for traffic flow based on wavelet analysis

Short-term forecasting for traffic flow is one of the key problems of traffic control and route guidance. With shortening of the forecasting horizon, the uncertainty of traffic flow becomes more and more serious, so that general approaches are incapable of forecasting the short-term volume effectively. In this paper, the feasibility of the short-term forecasting for traffic flow based on wavelet analysis is analyzed and an approach based on wavelet decomposition and reconstruction is put forward. Then the determination of the parameters is discussed and simulation results are given.