Photovoltaic parameters estimation using three metaheuristic algorithms: A comparative study

Sunlight has served as the primary energy source since the inception of life on Earth. Despite the emergence of alternative energy sources like fossil and nuclear energy, solar energy remains the most environmentally friendly and cost-effective option. Harnessing this energy involves utilizing photovoltaic (PV) modules to generate electricity. Extensive research is dedicated to PV modules, with a primary emphasis on electrical modeling, which plays a crucial role in effectively controlling a PV system and determining its I-V characteristics. PV modules encompass various electrical models, including the single-diode model (SDM), double-diode model (DDM), and triple-diode model (TDM). The difficulty lies in precisely determining the unknown parameters associated with each model. This study sets out with a clear objective: to tackle the challenge of identifying the elusive parameters within the SDM. The primary aim is to compare the effectiveness of three metaheuristic algorithms namely, the Flower Pollination Algorithm (FPA), Teaching-Learning-Based Optimization (TLBO), and Honey Badger Algorithm (HBA) in identifying these unknown parameters. In practical terms, this study extends to the evaluation of these algorithms on specific PV modules such as the Photowatt-PWP201 module, Tata Solar Power TP240 module, and RTC France solar cell. The evaluation of results is based on the root mean square error (RMSE) values. Notably, HBA stands out as it demonstrates superior performance, achieving the lowest RMSE of 9.860218e-04\;A for the RTC France solar cell. Conversely, FPA records the highest RMSE, reaching 9.458277e-03 A for the TP240 module.