Artificial neural network applied to prediction of electricity generated by Grid connected photovoltaic systems

This paper applies Artificial Neural Network to predict the amount of energy generated by a Grid Connected Photovoltaic System installed at the Institute of Electrotechnic and Energy of University of São Paulo (IEE/USP). Irradiance, back cell temperature and power data were collected during the period of one year. This methodology allows performing an analysis of the production of Grid Connected Photovoltaic Systems and the commercialization of the energy generated. Finally, the methodology was validated comparing relative error between measured data and estimated data.