An Hour-Ahead Photovoltaic Power Forecasting Based on LSTM Model

The extensive integration of the large-scale Photovoltaic (PV) plants into the power grid requires the development of new forecasting methods, for the prediction of the PV output with high accuracy. Despite the statistical and the Machine Learning (ML) approaches which have been extensively studied in the literature, the Deep Learning (DL) methods are not yet fully examined. Considering this, the present paper proposes a forecasting model based on the Long-Short Term Memory (LSTM) algorithm. Except of the solar irradiance, the module' temperature and the historical PV data, the influence of the clearness index to forecasting process has been also examined. The results indicate that the employment of the clearness index can improve the performance of the forecaster.