Using Limited-Memory Broyden-Fletcher and Emperor Penguin Algorithm To Minimize the Cell Voltage of Solid Oxide Fuel Cells

Abstract

Optimization methods are widely used in various industrial, scientific, and engineering applications to determine the most efficient planning strategy, determine the best distribution of a financial portfolio, design a logistics network in the most efficient way possible, or achieve the best performance of an artificial intelligence model. In this study, the aim is to minimize the cell voltage of solid oxide fuel cells to improve their performance and energy efficiency. In the optimization studies carried out with the L-BFGS-B algorithm and Emperor Penguin algorithm, the values of temperature (T), oxygen pressure (p(O-2 )), hydrogen pressure (p(H-2)), and water vapor pressure (p(H2O)) are calculated for minimum voltage while the input values of Faraday constant, Gas constant, Activation polarization coefficient, Reverse current density, and Electrode thickness are fixed. For both optimization methods, the optimum temperature value is calculated as 1000 K, the optimum oxygen pressure value as 1.0, the optimum hydrogen pressure value as 0.000001, and the optimum water vapor pressure value as 0.000001. The minimum cell voltage was calculated as 0.6486 for both optimization methods, but the L-BFGS-B algorithm reached the result in 7 iterations and 0.0046 seconds, while the Emperor Penguin algorithm reached it in 150 iterations and 1.18 seconds. According to the analysis results, although the cell voltage values of the two methods are the same, it can be seen that the L-BFGS-B algorithm is more successful in terms of iteration and time.