Efficient control of a nonlinear solid oxide fuel cell system using an adaptive model predictive controller

Abstract

Summary Over the years, the solid oxide fuel cell (SOFC) is growing commercially due to its high fuel flexibility, lesser maintenance requirement, and environmental friendliness feature. However, its control is challenging due to its nonlinear behavior and simultaneous management of its operational constraints. Therefore, in this work, an adaptive model predictive control is designed systematically to handle the SOFC's behavior while optimizing its operational constraints such as fuel input, utilization factor, and change in pressure of hydrogen and oxygen. This model‐oriented control linearizes, augments, and discretizes the SOFC during run time while estimating the states using a time‐varying Kalman filter. Finally, it optimizes the control problem for predicting the internal states of the SOFC using successive linearization. The simulation results show that the proposed controller gives better performance of the SOFC in linear region, and significantly improved performance in nonlinear region in comparison to the conventional model predictive controller. The validation of the work has been done through real‐time simulation using OPAL‐RT's hardware setup. Using such control is beneficial as it enhances the efficiency of the SOFC significantly. Thus, this advanced controller is most suited for these kinds of nonlinear systems due to its superior performance.