Traffic Flow Mixed Model Based on Wavelet Multi-resolution Analysis

The wavelet multi-resolution analysis is combined with time series traffic flow models. The characteristic of traffic flow is decomposed by wavelet multi-resolution analysis. According to the long relativity of smooth component and the short relativity of detail component, corresponding models are set up. The prediction of traffic flow series is implemented by synthesizing all the components. Experiments results indicate that the mixed model can make accurate prediction of traffic flow and the prediction results are better than using FARIMA model.