Short Term Prediction of Wind Power Generation Power Based on Variable Time Scales

The short-term prediction effect of wind power generation power is not entirely the same at different time scales. Considering that the annual statistical cycle cannot reflect the changes in wind power generation at different time scales, a short-term wind power prediction method based on variable time scales is proposed by combining the physical model of wind power generation and relevant historical data. By comparing the similarity based interpolation and linear interpolation power prediction methods through experiments, the superiority of the similarity based interpolation method in short-term power prediction is determined. On the basis of similarity interpolation research, short-term prediction of wind power generation based on variable time scales is carried out. The results show that this method can adaptively change the prediction cycle according to actual control requirements, and achieve variable time scale generation power prediction.