Restart and diversity-enhanced artificial rabbits optimization (RD-ARO) for photovoltaic parameter estimation
Year: 2026
Type: article
Abstract: Accurate photovoltaic equivalent-circuit parameter estimation is essential for reliable current–voltage and power–voltage modelling, but the optimisation landscape is highly nonlinear, multimodal, and prone to premature convergence. To address this issue, this study proposes Restart and Diversity-Enhanced Artificial Rabbits Optimisation (RD-ARO), an improved Artificial Rabbits Optimisation framework that preserves the original search kernel while introducing diversity monitoring, stagnation detection, adaptive partial restart, and elite preservation. The method was evaluated on four standard photovoltaic datasets, namely PVM752, RTC France, STM6, and STP6, under single-diode and double-diode model formulations. The results show that RD-ARO provides a stable and accurate estimation profile across both model types. According to Friedman mean ranks, RD-ARO achieved the best overall performance under both single-diode and double-diode models, with an average rank of 1.25 in each case. In the single-diode STM6 case, RD-ARO achieved a mean root mean square error of 1.979E-03 with a standard deviation of 2.125E-04. In the double-diode RTC France case, RD-ARO obtained a mean root mean square error of 7.479E-04 with a standard deviation of 8.770E-06. Additional ablation, sensitivity, runtime, convergence-speed, noise-robustness, and engineering operating-point analyses further confirmed that the improvement is mainly associated with the combined effects of stagnation monitoring, adaptive partial restart, and elite preservation. These findings indicate that RD-ARO is a competitive and stable optimisation framework for photovoltaic parameter estimation, especially when repeatability, robustness, and worst-run control are considered.
Source: Results in Engineering
Authors: Harun Gezici, Ercan Coşgun, Eray Yılmazlar
Institution: Kırklareli University
Language: English
Field: Energy
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $1,940
OpenAlex ID: W7171364298