Abstract

A method for modeling photovoltaic (PV) arrays - based on artificial intelligence techniques, namely genetic algorithm (GA) and cuckoo optimization algorithm (COA) - is presented. COA and GA are used to obtain the parameters of the PV array model using the PV cell’s datasheet information. The adopted models – using GA and COA - are implemented on a simulation platforms using MATLAB 2020a environment for two-diode and single-diode models. The proposed optimization method fits the mathematical current-voltage (I-V) characteristic to the three (V, I) remarkable points without the need to guess or to estimate any other parameter. The obtained models are tested and validated with experimental data taken from the Mutah university’s PV power plant. The results show that for both of the employed optimization algorithms, the two-diode model is more accurate than the single-diode model. The results also disclose that, at different values of temperature and solar irradiance, the COA – compared to the GA – better handles the optimization problem with low iterations and better fitness value.

Key words: Photovoltaic array; Photovoltaic cell; Genetic algorithm; Cuckoo optimization algorithm; Two-diode model; Single-diode model; Current-voltage characteristic.