A New Approach to Parameter Estimation in Direct Ethanol Fuel Cells Using an Extended Empirical Model

ABSTRACT In this study, a systematic framework is presented for the parametric characterization and performance analysis of direct ethanol fuel cells (DEFCs). An empirical polarization model from the literature is extended to explicitly include the effects of ethanol crossover and parasitic current density, resulting in a seven parameter voltage expression. For the optimization of the model parameters, a novel hybrid objective function is proposed that simultaneously minimizes the squared voltage error and the absolute power density error. The optimization is performed using the starfish optimization algorithm (SFOA) at two different temperatures, 25°C and 60°C. The obtained results demonstrate that the proposed model reproduces both the voltage and power density characteristics with high accuracy. Consequently, the total voltage error is calculated as 0.18 at 25°C and 0.58 at 60°C. The convergence curves and boxplots from multiple SFOA runs reveal that the algorithm produces consistent and well‐conditioned parameter sets. In addition, the conducted Sobol global sensitivity analysis shows that the variance of the objective function is primarily governed by the activation parameter and the effective Nernst potential, while the importance of the mass transfer coefficient increases at higher temperature. These findings indicate that the proposed modeling and optimization framework provides a reliable and computationally efficient tool for the design optimization of DEFCs, control oriented modeling, and their integration into various hybrid energy systems.