Wind power prediction model based on deep learning neural network

Wind power generation has strong randomness and volatility, and accurate prediction of wind power can improve the safety and reliability of grid operation. To further improve the accuracy of short-term wind power prediction, a long short-term memory (LSTM) short-term wind power prediction method is proposed. First, we selected a wind power plant, from 0:00 on July 1, 2022, to 22:38 on July 8, 2022, monitoring data of wind power generation per minute as the research object. Then, we Verify the rationality of dividing the data set according to the time series, and normalize the data set. The training set is used for training, and the training results are tested by the test set. Next, we established an LSTM prediction model, selectively forgotten and memorized information, and changed the way of time back propagation training, which largely overcame the problems of gradient disappearance and gradient explosion. Finally, compared to the Bagging class algorithm Random Forest and the Boosting class algorithm Adaboost, the average absolute percentage error (MAPE) of LSTM model is improved by 3.00% and 16.18%, respectively.it is concluded that the LSTM wind power prediction method has high prediction accuracy and strong applicability.