Work

Hybrid osprey-parrot optimization algorithm for enhanced parameter identification of supercapacitor model

Year: 2025

Type: article

Abstract: Supercapacitors (SCs) are important energy storage devices with widespread applications in various fields. These applications include renewable energy systems and electric vehicles. The ability of the SCs to deliver high power density, as well as their rapid charge-discharge cycles, makes them necessary. However, accurately modeling SCs remains challenging due to their complex dynamic behavior. A simple series RC circuit can approximate their characteristics. However, such a simple model fails to capture transient and long-term responses effectively. An advanced model incorporating three RC branches has been introduced to overcome this limitation. The first branch incorporates a capacitance that varies with voltage to account for immediate responses. The second branch models behavior within a short time frame. Meanwhile, the third branch accounts for long-term dynamics beyond 10 min. This comprehensive framework comprises eight distinct variables that should be optimally identified to achieve accurate performance prediction. A novel optimization approach is proposed to achieve accurate parameter identification by combining the Osprey Optimization Algorithm (OOA) and the Parrot Optimizer (PO). The hybrid approach has used the OOA's strong exploration capabilities in addition to the PO's efficient local search. This hybridization resulted in enhanced optimization performance. The proposed method is evaluated against other metaheuristic algorithms, including Genetic Algorithm (GA), Walrus Optimization Algorithm (WOA), and standalone OOA and PO. The performance is evaluated and compared based on the sum squared error (SSE) between experimental and simulated terminal voltage. The proposed hybrid approach achieves the lowest SSE value of 0.000173293. This confirms the superior accuracy. Statistical analysis further verified the robustness and reliability of the proposed hybrid algorithm. The strong correlation between model simulation results and experimental data highlights the effectiveness of the proposed hybrid optimization method for supercapacitor parameter identification.

Source: Journal of Energy Storage

DOI: 10.1016/j.est.2025.118604

Authors: Mohamed A.M. Shaheen, Hany M. Hasanien, Zia Ullah, Ahmad Alferidi, Mohammed Alsolami, Badr Lami

Institutions: Ain Shams University, University of Sharjah, Shanxi University, Taibah University

Language: English

Topic: Metaheuristic Optimization Algorithms Research

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

SDGs: Affordable and clean energy, Industry, innovation and infrastructure

Open Access status: hybrid

Estimated APC paid: $3,560