Parameters identification of photovoltaic models by using an enhanced adaptive butterfly optimization algorithm

Establishing accurate and reliable models based on the measured data for photo-voltaic (PV) modules are significant to design, control and evaluate the PV systems. Although many meta-heuristic algorithms have been proposed in the literature, achieving reliable, accurate and quick parameters identification for PV models is still a challenge. This paper develops a variant of butterfly optimization algorithm (called EABOA) to identify the unknown parameters of PV models. In EABOA, a new position search equation and good-point set are proposed to balance between exploration and exploitation. 12 classical benchmark test problems are firstly selected for verifying the effectiveness of EABOA, and the results indicate that EABOA provides better performance than other selected algorithms. Then, EABOA is applied to identify the unknown parameters of three benchmark test PV models, i.e., single diode (SD), double diode (DD) and PV module models. The comparison results with some other reported parameter identification methods from literature suggest that the proposed EABOA outperforms most approaches in terms of accuracy and reliability. The least SIAE value of EABOA is smaller than other compared algorithms about 56.6%, 5.84%, and 10.2% for SD, DD, and PV module models, respectively. Finally, EABOA is applied to solve parameter identification problem of practical module and obtains the satisfactory results. (c) 2021 Elsevier Ltd. All rights reserved.

Automated summary

The paper proposes a variant of butterfly optimization algorithm (EABOA) to identify unknown parameters of PV models, which shows better performance than other selected algorithms in terms of accuracy and reliability.