Artificial Intelligence Technics Applied to Analisys of Photovoltaic Energy Systems

This paper presents a procedure to determine the energetic behavior of photovoltaic systems through genetic algorithms and artificial neural networks. The method divides the I-V Curve of the photovoltaic generator in three regions. The best results are achieved for the region around the maximum power point, which can be advantageous for grid-connected systems. The method allows determining the energy delivered by the system and the inverter performance from a few experimental data. The output is a vector comprising the DC power, the maximum power point tracker efficiency, the inverter efficiency and the AC power.