Análise Comparativa de Modelos LSTM e GRU Para Previsão da Velocidade do Vento.

This work aims to compare the prediction performance of the LSTM (Long Short Term Memory) and GRU (Gated Recurrent Units) models. To carry out this comparison, a database of maximum wind speed, obtained from the National Institute of Meteorology (INMET), was used. This database presents a monthly historical series of maximum wind speed, from the Palmeira dos Índios meteorological station, in the period between 2008 and 2020. The LSTM and GRU forecast models were implemented in Python using the Pytorch library. Results obtained from the two models were compared using the metrics RMSE (Root Mean Squared Error), MAPE (Mean Absolute Percent Error) and MAE (Mean Absolute Error). It was verified, for a short-term horizon (6 months), that the LSTM neural network presented the best performance. [ABSTRACT FROM AUTHOR]