Model identification of Solid Oxide Fuel Cell using hybrid Elman Neural Network/Quantum Pathfinder algorithm

In this research, a new efficient method is introduced for model assessment of Solid Oxide Fuel Cell (SOFC) model using a new hybrid Elman Neural Network (ENN). The main purpose of this research is to minimize the Mean Squared Error (MSE) between empirical data and modeling data of the fuel cell output voltage using the suggested hybrid ENN. The designed ENN is indeed a combination of this network with an improved metaheuristic, called Quantum Pathfinder (QPF) algorithm to give an optimal model. The proposed QPF-ENN model is then performed in a SOFC case study to show its efficiency. The results of the suggested method are validated by the reference voltage and also two other methods to show the higher minimum value of the Mean Squared Error (MSE) toward the others. Simulation results are analyzed the mean squared error value of the methods for 5000 samples, where, the voltage is limited between 320 V and 361 V. The results show that the mean square error for the QPF-Elman method, GWO-RHNN method, and PF-Elman method are 0.0014, 0.0017, and 0.0018, respectively. This indicates that the proposed QPF-Elman delivers the minimum value of the mean square error.