Short-Term Wind Power Prediction Based on AVMD-SMA-LSSVM Combined Model

This research proposes a combined wind power prediction model to increase wind power forecast accuracy and energy consumption efficiency. To begin, the adaptive variational mode decomposition (AVMD) method is used to deconstruct the wind power signal at various scales and create a number of sub sequences. Second, a prediction model for subsequences is built using the slime mold algorithm (SMA) and least squares support vector machine (LSSVM) parameters that are adaptively determined. Finally, the subsequence prediction values are weighted and fused to get the final wind power forecast value. Using the AVMD approach to minimize non-stationary and noise interference in signals and increase subsequence prediction accuracy. Developing a novel intelligent optimization technique that combines SMA and LSSVM to eliminate mistakes caused by human parameter setup. Weighted fusion is used to subsequence prediction results, completely using the prediction information of each subsequence to increase the stability and reliability of the prediction outcomes. The proposed model has greater prediction accuracy and reduced prediction error, as demonstrated by example analysis in this article.