Refined power modeling method of ultra-short-term wind power cluster based on power reverse deduction

The accurate prediction of wind power provides a guarantee for the safety and stability of the power system and the power market trading. The existing wind power prediction methods mainly focus on the depth of the model, and improve the accuracy of the prediction by constantly improving the model and deepening the model, but they ignore the limitations of the model itself. In this paper, a fine modeling method for wind power is proposed, and the prediction accuracy is improved by horizontal expansion of the model. Firstly, the power output class and its NWP characteristics are determined by power reverse deduction, and the power output class is divided by the degraded and simplified U-net model. Then, Non-stationary Transformer (NSTransformer) is introduced to model each output level separately to get the predicted power. Finally, the proposed method is applied to a wind farm cluster in China. The accuracy of the proposed method is higher than other methods in ultra-short-term prediction, which proves the reliability and effectiveness of the proposed method.