First, a single LSTM model was built to predict the wind power. Then a CNN-LSTM combination model was built. It is shown that the combination model could achieve better performance(measured by RMSE/MAE).
The results are as follows:
(1)Predict in 6h
(2)Predict in 12h
(3)Predict in 24h
(4)Predict in 48h
(5)Predict in 96h
(6)Predict in 192h
And RMSE/MAE is as follows:
jlian2/Tensorflow-Wind-Power-Prediction
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|