Development of a Machine Learning Algorithm to Forecast PV Plant Production

The prediction of photovoltaic (PV) energy generation is crucial to enhance the efficiency of renewable energy systems. This paper proposes the use of machine learning models such as Random Forest to forecast photovoltaic energy production, focusing on integrating environmental variables such as temperature, humidity, wind speed. The Random Forest model was implemented to explore the influence of temporal and meteorological variables. The results indicate that the combined use of the advanced machine learning techniques led to much more accurate forecasts measured by Mean Squared Error (MSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE), contributing to improving the stability of the electrical system by reducing imbalances. The results obtained from the analysis of historical and meteorological data from a real photovoltaic system show significant potential for their use in optimal management models, both for individual users and renewable energy communities.