An Optimal Model Identification for Solid Oxide Fuel Cell Based on Extreme Learning Machines Optimized by Improved Red Fox Optimization Algorithm

The present study proposes an efficient method for optimal model parameters estimation of the Solid Oxide Fuel Cell by considering its nonlinear dynamic behavior. The approach is relied on using a novel optimal model of Extreme Learning Machines (ELM) network based on metaheuristics. The main purpose is to minimize the Mean Squared Error (MSE) between the empirical output voltage data and the output voltage of the model by the suggested optimized ELM network. The proposed ELM network is optimized by an enhanced design of the Red Fox Optimizer (IRFO) Algorithm to provide optimal results. The suggested ELM-IRFO method is then testified on a Solid Oxide Fuel Cell case study and its results are compared with the GWO-RHNN method from the literature and ELM-RFO method to show its effectiveness. The final results showed that the proposed ELM-IRFO has the minimum value of the Mean Squared Error (MSE) against the other comparative methods.