Chaotic characteristics analysis and prediction model study on wind power time series

Abstract

In order to reveal the internal dynamic property of wind power time series the nonlinear analysis method is used to identify the chaotic property of wind power set which is the basis for the prediction of the wind power time series. Firstly day correlation property on wind power time series of a certain wind farmer is analyzed. Secondly the largest Lyapunov exponent of wind power set is calculated on the basis of phase space construction to verify the presence of chaos in wind power time series. The ultra-short-term predicted of wind power would produce larger errors by using the Volterra filter multi-step prediction so the predicted results of Volterra filter are corrected by combining the results predicted by Local-region Multi-steps Method and the largest Lyapunov exponent method with weighted Markov chain and ordered operator. Finally the prediction on wind power of a certain wind farmer is presented and the simulation results illustrate that the correction forecasting model improves high predictive accuracy effectively, which provides a useful reference for wind power prediction by the Volterra filter multi-step method.