Optimization of Power Output Forecasting from Solar PV System using Deep Learning Methods

This research proposes a deep learning method (GA-RNN-LSTM) for forecasting the power output from a solar PV system. The model takes six input parameters, and the output is the power output from the PV system. The proposed model was compared to the conventional RNN-LSTM model in terms of accuracy. The evaluation metrics used were Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Coefficient of Determination (R2). The results demonstrate that the GA-RNN-LSTM model demonstrates superior forecasting accuracy and robustness compared to the RNN-LSTM model. Specifically, the proposed model (GA-RNN-LSTM) obtained the best testing results with an RMSE and MAE of 2.581 and 2.065, respectively