Optimal parameter identification of SOFC model using modified gray wolf optimization algorithm

Abstract

Industrial and commercial use of fuel cells as a source of clean energy production are two important goals of researchers in the field of energy today. Therefore, researchers in this science are always looking for new methods for industrial production and at a reasonable price for fuel cells. This study presents a new well-organized methodology for model identification of a Solid Oxide Fuel Cell (SOFC) stack by providing optimal selection of the unknown variables in the model. The main objective here is to minimize the sum of squared error value between the designed model output voltage and the experimental data. Here, to provide an optimization process, a modified version of gray wolf optimization (MGWO) algorithm has been utilized. This algorithm is then utilized to improve the algorithm efficiency and to get better results. To show the system reliability, two scenarios based on temperature and pressure variations have been utilized. The technique has been finally compared with several other techniques to verify its prominence. With considering the achieved results, it can be observed that the sum of square error for different temperature values based on the proposed method is too small, such that for the temperatures with values 553.4 °C, 652.3 °C, 669.8 °C, 754.6 °C, and 800 °C the SSE value is 5.27 e−4, 2.66 e−4, 3.91 e−6, 4.19 e−3, 2.07 e−4, respectively. Furthermore, the pressure value variations from 1 atm to 5 atm with 1.44 e−3, 3.20 e−3, 5.84 e−3, 3.36 e−3, 2.67 e−3, respectively indicate its higher efficiency toward the other studied methods. Final results designate that the proposed technique delivers outstanding efficiency toward the compared methods.