Study of the short-term traffic flow forecasting based on nonlinear theory
Qixin Shi · Journal of Xi'an University of Architecture & Technology · 2006
This paper tries to analyze and forecast the short-term traffic flow based on the nonlinear theory.The traffic flow forecasting frame is presented first and Three kinds of traffic flow data which are collected in different time scales(1 min,5 min and 15 min) are analyzed using PCA method.It shows that the traffic flow data exhibits chaotic properties.On this basis,the first-order weighing local approximation method based on phase space reconstruction is set up.At last,the paper applies the forecasting models to the real traffic flow time series in Shanghai viaduct of Yan'an road.The result of is forecasting satisfying.