The Hybrid Tianji’s Horse Racing Optimization–Enzyme Action Optimizer: A New Hybrid Algorithm for Parameter Estimation of Lithium-ion Battery Model
Arabian Journal for Science and Engineering, 2026 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Basım Tarihi: 2026
- Doi Numarası: 10.1007/s13369-026-11690-4
- Dergi Adı: Arabian Journal for Science and Engineering
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, zbMATH, Materials Science & Engineering Collection (ProQuest), Technology Collection (ProQuest)
- Anahtar Kelimeler: Enzyme action optimizer, First-order and second-order RC models, Hybrid algorithm, LiB parameter estimation, Lithium-ion battery, Tianji’s horse racing optimization
- Gazi Üniversitesi Adresli: Evet
Özet
This article presents a new hybrid metaheuristic algorithm, the hybrid Tianji’s horse racing optimization (THRO)–enzyme action optimizer (EAO). The proposed hTHRO-EAO has been developed based on operating the THRO and EAO in a uniquely cooperative manner. It is proposed for the joint nonlinear optimization of the open-circuit voltage–state-of-charge (OCV–SoC) relationship and the parameters of lithium-ion battery (LiB) equivalent-circuit models. A sixth-order OCV–SoC polynomial, charge- and discharge-dependent circuit parameters are simultaneously identified for first- and second-order RC models. Accurate determination of these parameters is crucial for reliably representing the electrical behavior of batteries. 14 algorithms are evaluated under 6 optimization scenarios using DST datasets at 0, 25, and 45 °C and 50% SoC. All algorithms are tested under identical parameter bounds, initial conditions, population sizes, and iteration limits over 30 independent runs. Across the investigated scenarios, hTHRO-EAO demonstrates strong performance in terms of minimum, maximum, average, and standard deviation RMSE values (6.68306E−03 V, 7.49653E−03 V, 7.18144E−03 V, and 3.5865E−05 V, respectively). Furthermore, statistical analyses confirm the superiority of the proposed algorithm. Friedman and Wilcoxon tests confirm the statistical superiority of hTHRO-EAO, which achieves the best average ranking in both model structures. Cross-profile validation is performed by transferring the DST-identified second-order RC parameter sets to previously unseen BJDST and FUDS profiles without reoptimization. The validation RMSE values range from 1.0194E−02 to 4.1580E−02 V. Overall, hTHRO-EAO provides an accurate, stable, and transferable framework for joint OCV–SoC and equivalent-circuit parameter estimation.