MEJORA DE LA EFICIENCIA ENERGÉTICA DE UNA PLANTA FOTOVOLTÁICA CON EL USO DE REDES NEURONALES ARTIFICIALES.

Abstract

The scope of this project is obtaining an ideal electrical power generation model based on artificial neural networks with back-propagation, of a 6,1 MWp photovoltaic power plant located in Cordoba (Spain) operating since september of 2008. The aforementioned model is used with several objectives focusing in efficient energy improvement. With this aim in mind, a critical analysis and warning settlement policy in a SCADA system are developed, in a way that it is possible to anticipate the failure event, comparing the output of the model with the actual power generation. It has been tested the methodology on an specific failure mode to analyze the goodness of the model, reaching a 1,8 % improvement in electrical power production.