Photovoltaic Power Generation Forecasting Utilizing Long Short Term Memory

The modernization of the world has considerably reduced the prime sources of energy, such as coal, diesel and gas. Thus, alternative energysources based on renewable energy have been the main concentration nowadays to address the world’s energy demand and at the same time torestrict global warming. Among these renewable energies, solar energy is the main source used to generate electricity through photovoltaic(PV) systems. However, the output of PV power is highly intermittent. Thus accurately forecasting the output of PV systems is an importantrequirement to ensure the stability and reliability of the grid. This study develops and validates a short-term PV power forecasting modelby using the combination of a genetic algorithm (GA) and Long Short Term Memory (LSTM). The performance of the proposed model iscompared with LSTM baseline model by three errors (Root mean square error(RMSE), Mean absolute error (MAE), and Mean absolutepercentage error (MAPE)) in two case studies.