Research on Wind Power Prediction Combined with Prediction Models and the Improved Salp Algorithm

Because of the fluctuation of wind power, it is very important to regulate it by means of prediction. After a comprehensive analysis of the current research status in this field, this paper proposes the decomposition method of the singular spectrum decomposition as the data set, and then uses the improved salp algorithm to complete the combined optimization of the four prediction models of recurrent neural network, long and short term memory neural network, gating cycle unit and Extreme Gradient Boosting, so as to realize the wind power prediction. Finally, through experiments on the wind power data set derived from the wind farm in Yantai, China, Shandong Province, China, the value of R2 of this model reached 0.9968, indicating a high degree of fit to the wind power data. The model proposed in this paper can accurately complete wind power prediction, thereby effectively reducing the impact caused by the volatility of wind power generation.