A nature-inspired evolutionary algorithm, black widow optimization (BWO), for parameter identification is proposed and applied to a 5-kW tubular solid oxide fuel cell system. A series of test runs on minimization of standard benchmark functions have been performed and compared with other metaheuristic algorithms. In terms of accuracy, robustness, convergence, and statistics, BWO is a competitive method.

Abstract

A new effective method for optimization of unknown parameters in solid oxide fuel cell (SOFC) stack models is suggested. The overall voltage of the SOFC stack depends on these predicted parameters. The goal is to reduce the mean square error (MSE) between the empirical and the predicted polarization curve obtained using the method. Black widow optimization (BWO), a metaheuristic method inspired by nature, describes the minimization process. This algorithm is made to alter the search space, avoid local optima, and deliver greater efficiency in the exploitation and exploration stages. Situations based on multidimensional benchmark functions and SOFC stack temperature variations are investigated to ascertain the system consistency.