Computational modeling of a 119 kWp photovoltaic system

An important aspect of medium and large photovoltaic systems is the ability to predict future electricity generation based on predicted meteorological data as well as the technical and structural characteristics of the system. This prediction can be enabled by applying mathematical models or artificial intelligence. This study aims to develop a reliable and valid computational model using measurements from a real 119 kWp photovoltaic plant at the Federal Institute of Espirito Santo. The model considers meteorological data from a solarimetric station located in the same area, as well as system losses in both AC and DC components. The results indicate that the simulated generation closely aligned with the actual figures, achieving an acceptable low percentage error, less than 0.5%, during the period from September to December 2022.