Research on Wind Speed Prediction Model Based on WOA-LSTM

Abstract

To tackle the issues of haphazardness of LSTM model boundaries and slow model intermingling speed, work on the exactness of wind speed forecast model, and better work on the security and economy of wind power age frameworks, this paper proposes an improved LSTM based on the Whale Optimization Algorithm (WOA), and conducts simulation analysis based on the US meteorological data. The exploration results show that:1) Through persistent preparation to track down reasonable hyperparameters and confirm them, it is observed that the attributes of the WOA calculation can successfully track down the worldwide ideal answer for upgrade the boundaries of the LSTM model; 2) Compared with the traditional LSTM neural network prediction model, the WOA-LSTM reduces the root mean square error (RMSE) and mean absolute error (MAE) by 0.43035 and 0.23115, respectively, and improves the goodness of fit (R^2) by 0.03763. It shows that the model has better accuracy and stability for wind speed prediction.