Investigation of a social group assisted differential evolution for the optimal PV parameter extraction of standard and modified diode models

Abstract

Parameter estimation of PV cells is essential to develop accurate simulation techniques, optimal control mechanisms, converter design optimization, maximize the efficiency of PV harvesting systems, improve the accuracy of monitoring systems etc. Advanced meta-heuristic algorithms for the optimal parameter estimation through the minimization of the disparity between the measured and estimated I-V data pairs have emerged as a powerful tool in this regard. Despite the efficacy of the meta-heuristics for the standard diode models, the optimization of modified diode models including the higher-order complex modified double and triple diode modules remains largely unexplored. Hence, an advanced hybrid meta-heuristic technique benefitting from the synergy of social group optimization and differential evolution known as social group assisted differential evolution (SGDE) is developed to achieve an enhanced balance of exploration and exploitation for the complex multi-modal landscapes of the modified diode models. Three experiments involving the minimization of the root mean square error for three commercial PV modules (KC200GT, Shell SQ85 and Shell ST40) based on the simulated I-V data followed by four experimental I-V curves of two solar cells (RTC France solar cell and PVM 752 GaAs thin-film cell) and two PV modules (Photowatt-PWP201 and STM6-40/36 PV modules) and the parameter optimization for two commercial modules (JAM72S01-380/PR and SUNCECO SEP300W PV modules) at varying levels of irradiance and temperature are investigated and extensively compared with fourteen other state-of-the-art meta-heuristic techniques. SGDE outperforms the competitor algorithms in terms of optimality, consistency and demonstrated greater robustness to local entrapment for the standard and modified diode models.