Wind Speed Prediction Based on Improved Empirical Mode Decomposition and Hybrid Deep Learning Models
Accurate wind speed prediction is of great significance for wind power consumption and the stable operation of power system.This paper proposes to combine the improved complete ensemble empirical mode decomposition with adaptive noise(ICEEMDAN)method with a hybrid deep learning model to improve the accuracy of wind speed prediction.Firstly,ICEEMDAN is used to extract different frequency features in complex wind speed sequences.Then a temporal convolution network(TCN)-gated recurrent unit neural network(GRU)model is constructed for the different frequency features to obtain the long-term time-series information and predict the sequence of each feature.Finally,the predicted value of each feature sequence is weighted and integrated as the final result.The experimental results show that the proposed ICEEMDAN-TCN-GRU model has high prediction accuracy and strong stability compared with the contrast model.
wind powerwind speed predictiontime series decompositionTCN networkGRU neural networks
NETL
NSTL
万方数据
维普
