A Hybrid Deep Learning-Based Network for Photovoltaic Power Forecasting

Figure 1

The proposed PV power forecasting framework comprises mainly three steps. Step 1: Preprocessing is applied to normalize the data, remove outliers, and fill in missing values. Step 2: Model selection with the refined data using the GRU and CNN networks. Step3: Evaluate the model via metrics, including MSE, MAE, RMSE, and MBE.