Estimation of the parameters of the mathematical model of an equivalent diode of a photovoltaic panel using a continuous genetic algorithm

• Published in: Revista IEEE América Latina Vol. 20; no. 4; pp. 616 - 623
• Main Authors: Montano, Jhon Jairo, Grisales Norena, Luis F., Tobon, Andres Felipe, Gonzalez Montoya, Daniel
• Format: Journal Article
• Language: English; Spanish
• Published: Los Alamitos IEEE 01.04.2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Documents; Estimation; Genetic algorithms; Integrated circuit modeling; Mathematical model; Mathematical models; modeling; Optimization methods; Parameter estimation; parameters identification; Particle swarm optimization; photovoltaic; Photovoltaic cells; Photovoltaic systems; Renewable energy; Resistance; Resistance factors; Robustness (mathematics); Semiconductor diodes; Solar
• ISSN: 1548-0992; 1548-0992
• DOI: 10.1109/TLA.2022.9675467

This document presents the implementation of a con- tinuous population genetic optimization algorithm (CGA) as a solution method to the parameter estimation problem of a diode model (SDM) of a photovoltaic panel (PV) from experimental data of voltage versus current (V-I). The parameters to be estimated by means of the CGA are: the photoinduced current, the diode saturation current, the ideality factor, the series resistance and the parallel resistance. The estimation of the SDM parameters is carried out in order to obtain the real values that represent the power profile of the panel and thus carry out an analysis of its physical state. For which, the mean square error of the PV current estimated by the solution method from the selected parameters is used as the objective function, with the real curve of the PV panel used as the test scenario. All of the above subject to the set of restrictions that limits the problem under analysis. To validate the effectiveness and robustness of the proposed method, in this document two comparison methods have been used: the particle swarm optimization method (PSO) and a traditional genetic algorithm (GAT). In addition, four different panel types were used to generate the test scenarios: the MSX60, the SOLAR SJ65, the KYOCERA KC200GT, and the STP245S. All simulations were obtained using MATLAB 2019b. The results obtained in this document show that the proposed method presents the best relationship between the estimation of parameters and the computation time required to solve the SDM problem.