Wind Power Prediction Method: Support Vector Regression Optimized by Improved Jellyfish Search Algorithm

To address the problems of grid connection and power dispatching caused by non-stationary wind power output, an improved Jellyfish Search algorithm optimization support vector regression (IJS-SVR) model was proposed in this study to achieve high-precision wind power prediction. The random selection of internal parameters of SVR model will affect its performance. In this study, the Jellyfish Search (JS) algorithm was selected and improved to propose an Improved Jellyfish Search (IJS) algorithm. Compared with the comparative algorithms, the optimized values of IJS algorithm are closer to 0. It exhibits good convergence ability, search stability, and optimization-seeking ability, as well as being more suitable for solving optimization problems. Therefore, IJS was used to optimize SVR, and the prediction model of IJS-SVR was established. Different weather and seasons affect wind power and model prediction accuracy. The wind power in spring and winter was selected for model prediction verification in this study. Compared with other methods, the IJS-SVR model proposed in this study could achieve better prediction results than other models in both seasons, and its prediction performance was better, which could improve the prediction accuracy of wind power. This study provides a more economical and effective method of wind power to solve its uncertainties and can be used as a reference for grid power generation planning and power system economic dispatch.

Automated summary

An improved Jellyfish Search algorithm optimization support vector regression (IJS-SVR) model is proposed in this study to achieve high-precision wind power prediction. The model shows better prediction results in spring and winter compared to other models, improving the prediction accuracy of wind power. This study provides an economical and effective method to address the uncertainties of wind power.