Improvement of Time-Series Method for Short-Term Traffic Prediction
Yuan Hong-kai · Journal of Xuzhou Institute of Architectural Technology · 2005
The time-series method has come into wide use in short-term traffic prediction.This paper analys the merits and demerits of the method.On the basis of the traditional time-series method,an improved method is obtained by inducting the recursive forgetting factor least square method(RFFLS),which brings about the real-time online prediction and smoothes away the problem of random severe interferences and great uncertainty in the real-time prediction of short-term traffic flow.