Parameter extraction of photovoltaic cells and modules by hybrid white shark optimizer and artificial rabbits optimization

ÇETİNBAŞ İ., TAMYÜREK B., DEMİRTAŞ M.

Energy Conversion and Management, cilt.296, 2023 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 296
  • Basım Tarihi: 2023
  • Doi Numarası: 10.1016/j.enconman.2023.117621
  • Dergi Adı: Energy Conversion and Management
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, PASCAL, Aerospace Database, Applied Science & Technology Source, CAB Abstracts, Communication Abstracts, Compendex, Computer & Applied Sciences, Environment Index, INSPEC, Pollution Abstracts, Veterinary Science Database, Civil Engineering Abstracts
  • Anahtar Kelimeler: Artificial rabbits optimization, Double diode model, Hybrid algorithm, PV parameter extraction, Single diode model, White shark optimizer
  • Eskişehir Osmangazi Üniversitesi Adresli: Evet

Özet

This study presents the parameter extraction of photovoltaic (PV) cells and modules using a new hybrid metaheuristic algorithm developed based on the white shark optimizer (WSO) and the artificial rabbits optimization (ARO) algorithm. The accurate forecasting of economic benefits and the performance of PV based applications are heavily dependent upon the model accuracy of PV devices. Therefore, the main objective of this study is to obtain the most accurate model for the PV cells and modules that are commercially available in the market. For this purpose, the proposed hybrid algorithm (hWSO-ARO) is used for the modeling of two PV cells and six PV modules made by different manufacturers. A total of sixteen models for eight PV devices consisting of single diode (SDM) and double diode model (DDM) for each device have been developed. To verify the success, the results are compared against the results of seven different state-of-the-art algorithms. According to the results of 30 runs, the proposed algorithm has yielded the smallest objective function/root mean square error (RMSE) values in average, maximum, and minimum evaluation metrics. The hybrid algorithm achieves the best minimum average RMSE of 1.8354513E−04. It is then followed by the WSO algorithm with 1.8557680E−04 and with a difference of 1.11% and the ARO algorithm with 2.1286641E−04 and with a difference of 15.97%. Similarly, when the maximum RMSE metric is evaluated, the hybrid algorithm comes the first again with the lowest maximum RMSE of 2.0001806E−04. It is then followed by the WSO with 2.0897277E−04 and with a difference of 4.48% and again the ARO with 2.2591014E−04 and with a difference of 12.94%. When the last metric, the standard deviation, is analyzed, hWSO-ARO, out of 16 models, achieves the most minimum standard deviation of 3.6904635E−19. In addition, the Friedman rank test issues the highest rank to the hybrid algorithm. Moreover, the Wilcoxon rank test is producing results in favor of the proposed algorithm in 110 of 112 pairwise comparisons of algorithms and showing a 98.21% success rate. Later, a sensitivity analysis is performed using the datasheets values of the selected cells and modules. At every comparison, a perfect match is achieved between the calculated and the experimentally measured data. It is even maintained under variable irradiance and temperature conditions. In conclusion, it is demonstrated that the hybrid algorithm stands out in solving the PV parameter extraction problem with great success under all conditions and enables the development of the most accurate PV models with high computational accuracy.