Short-term Traffic Flow Forecasting based on the Improved Non-parametric Regression

In ITS field, short-term traffic flow forecasting is one of the key technologies for traffic control and guidance. One kind of short-term traffic flow prediction method based on the improved non-parametric regression model is proposed in this paper. In this method of distance metric criteria, the Dynamic time warping distance is used instead of the traditional Euclidean distance as the prediction method. Using the non-linear alignment of each point between two sequences to calculate the similar distance, it can overcome the matching problem caused by the expansion and contraction of time series in timeline, and get a better forecast result. The performed simulation based on the traffic data of Xiamen Lotus junction cross-section shows the lower prediction errors that indicates the feasibility of this method.