Solar photovoltaic model parameter estimation based on orthogonally-adapted gradient-based optimization

Abstract

Solar photovoltaic (PV) model parameter estimation research is a growing field of interest. To establish accurate and reliable PV models, including single diode, double diode, three diode and PV module models, a new method is proposed in this paper. Gradient-based optimization (GBO) is designed to be used in conjunction with orthogonal learning, namely, orthogonal learning gradient-based optimization (OLGBO). In OLGBO, the orthogonal learning (OL) mechanism improves the speed rate and accuracy of GBO, achieving an adaptive conversion between exploitation and exploration. Therefore, the proposed OLGBO is evolved into a way for unknown parameter estimation of different PV models. Compared with the RLGBO and basic GBO methods, OLGBO achieves a smaller root-mean-square error (RMSE) and standard deviation (STD). In addition, OLGBO is compared with other algorithms, and the outcomes also demonstrate that the OLGBO method has a better excellent effect than selected state-of-the-art methods at certain temperatures and light conditions. Thus, all the evidences indicates that OLGBO can be a fast, promising, reliable, and feasible optimization method for dealing with unknown parameter identification problems in photovoltaic models.