Wind Power Prediction Method Based on CEEMDAN Decomposition
To address the issue of low prediction accuracy in wind power forecasting caused by the randomness and volatility in wind energy,a wind power prediction method based on CEEMDAN is proposed.Firstly,the CEEMDAN algorithm is used to decompose the wind power signal.Then,utilizing the LSTM algorithm as the foundation,an ELDGAWP wind power prediction model is designed by incorporating the Encoder-Decoder framework and Attention mechanism.This effectively solves the problem of gradient vanishing that LSTM models may encounter when dealing with very long sequence data.Finally,the predicted results of each mode component are accumulated to obtain the final prediction result.Compared with existing models,the proposed C-ELDGAWP prediction method achieves the highest prediction accuracy.
万方数据
维普
