A High-Efficiency Satellite-Based Optimization Approach for the Characterization of Solid Oxide Fuel Cell (SOFC) Models

• A new, ASSA, is introduced for robust SOFC parameter identification and ASSA inspired by orbital mechanics. • ASSA is validated on ten SOFC simulation cases covering realistic thermal and pressure conditions and • Comprehensive benchmarking against nine state-of-the-art algorithms proves ASSA’s superiority in both performance and stability. • ASSA’s achieves MSE as low as 10⁻⁶, 82-99% improvement over nine state-of-the-art algorithms and Computational speed of 0.113 seconds—237 × faster than competing methods. • Validated across ten operating conditions and experimental data with R² > 0.99. The solid oxide fuel cell (SOFC) stands as a vital clean energy technology; however, accurate parameter identification remains challenging due to its nonlinear, multi-physical characteristics. This study introduces the Artificial Satellite Search Algorithm (ASSA), a novel physics-inspired metaheuristic that mimics satellite orbital mechanics employing Medium Earth Orbit (MEO) for global exploration and Low Earth Orbit (LEO) for local exploitation to identify seven unknown parameters (E₀, A, I₀, c , R ohm , B, I_L, I₀,ₐ) in SOFC electrochemical models. ASSA demonstrated mean squared errors (MSE) as low as 1.6810 -5 (at 1123K) and 5.510 -6 (at 9 atm) as a 8299 average compared to 9 state-of-the-art competitors including SDO, SSA, WOA, and GWO, and was also validated at ten operating conditions between temperatures of 1073K and pressure of 1 atm down to 9 atm. ASSA showed standard deviation 40-85 times smaller than the competitor algorithms, and its calculations took an average of 0.113 seconds 237 times faster than the slowest competitor (SCSO) and 42 times faster than GWO. Association Statistical validation using Friedman Ranking Tests placed ASSA first in all cases (assigned average score 1.4), and convergence curves and box plot analysis verified that it would converge quickly and be robust. The parameters found gave V-I and P-V curves that corresponded to simulated values with R2 of greater than 0.999 in most running conditions, indicating that ASSA can be used to model, control and diagnostics SOFC in a real time and with high accuracy.