A novel wind speed forecasting based on hybrid decomposition and online sequential outlier robust extreme learning machine

As the wind energy developing, wind speed prediction is important for the reliability of wind power system and the integration of wind energy into the power network. This paper proposed a novel model based on hybrid mode decomposition (HMD) method and online sequential outlier robust extreme learning machine (OSORELM) for short-term wind speed prediction. In data pre-processing period, wind speed is deeply decomposed by HMD, which is comprised of variational mode decomposition (VMD), sample entropy (SE) and wavelet packet decomposition (WPD). The crisscross algorithm (CSO) is applied to optimize the input-weights and hidden layer biases for OSORELM, which have impact on the forecasting performance. The experiment results show that: (a) HMD is an effective way of wind speed decomposition, which can capture the characteristics of wind speed time series accurately and thus promote the prediction performance; (b) the OSORELM performs better than offline models in practical forecasting; (c) the proposed forecasting model has greatly improved the accuracy in mult-istep wind speed forecasting.