Short-term wind power probabilistic forecasting based on SSA-VMD-LSTM-NKDE

Abstract

In order to further improve the accuracy of wind power forecasting, a combined forecasting method based on sparrow search algorithm (SSA) optimizing variational mode decomposition (VMD) parameters was proposed. Firstly, the SSA was used to optimize the VMD parameters, then the optimized VMD was used to decompose the data. Secondly, the entropy weight method and grey relational analysis were combined to analyze the correlation of environmental variables, and the combination of the most relevant influencing factors and the decomposed modal components were selected as the input of the LSTM prediction model to obtain more accurate prediction results. Finally, a wind power probability prediction model based on NKDE was established to effectively quantify the uncertainty of wind power prediction results. The results show that the MAE, RMSE and MAPE of the proposed combination model decrease by 3951%, 33.22% and 40.39%, respectively, compared with the VMD-LSTM model. The SSA-VMD-LSTM-NKDE combination model can not only effectively improve the accuracy of deterministic prediction, but also effectively quantify the uncertainty of wind power prediction results, which provides scientific decision-making basis for wind power prediction.

Keywords

• wind energy; sparrow search algorithm; variational mode decomposition; entropy weight method; grey relational analysis; combined prediction mode