Robust parameter estimation in solid oxide fuel cells using a multi strategy improved crayfish optimization algorithm

This study addresses the challenge of accurate parameter estimation in Solid Oxide Fuel Cells (SOFCs), which is vital for precise modeling and performance optimization under varying thermal and pressure conditions. Conventional estimation techniques and existing metaheuristic algorithms often suffer from local optima entrapment, instability, or infeasible predictions in nonlinear SOFC models. To overcome these challenges, we propose the Multi-strategy Improved Crayfish Optimization Algorithm (MICOA), an enhanced nature-inspired framework integrating cave selection, food attraction, and Cauchy mutation strategies. MICOA was validated across ten SOFC operating scenarios (five temperatures: 1073–1273 K; five pressures: 1–9 atm) and benchmarked against nine state-of-the-art algorithms, including Tunicate Swarm Algorithm (TSA), Modified Sand Cat Swarm Optimization Algorithm (MSCSO), African Vultures Optimization Algorithm (AVOA), Duck Swarm Algorithm (DSA), Crayfish Optimization Algorithm (COA), Sine Cosine Algorithm (SCA), Arithmetic Optimization Algorithm (AOA), Reptile Search Algorithm (RSA), and Whale Optimization Algorithm (WOA). Quantitative analysis shows that MICOA consistently achieved the highest estimation accuracy, with minimum mean squared error values as low as , mean errors below 0.05 in most cases, and standard deviations under 0.04, confirming stability. Computational efficiency was equally strong, with runtimes of 0.10–0.18 seconds—over 40 times faster than RSA and more than 300 times faster than TSA. Friedman ranking tests further established MICOA’s dominance, with an average rank of about 1.1 across all cases. These findings confirm both the physical relevance and practical applicability of MICOA, positioning it as a promising tool for digital twin frameworks, fault diagnosis, and real-time SOFC optimization.