An Adaptive Kalman Filter for Short-Term Traffic Flow Forecasting

The paper presents an adaptive model based on the Kalman Filter Model (AKFM) for short-term traffic flow forecasting. Simultaneously, it expounds the basic principles and the implementation process of AKFM in detail. In addition, the paper has implemented AKFM and Classical Kalman Filter Model(CKFM) in C++ and evaluated them by three kinds of different ways. Experimental results show that AKFM integrates the advantages of the adaptive method, which can automatically adjust the parameters according to the model situations. Furthermore, the stability of AKFM is the better than CKFM. It can promote the prediction efficiency and lower the relative error of prediction in limited time. In a word, AKFM is effective and stable.