Model Predictive Control for Power and Thermal Management of an Integrated Solid Oxide Fuel Cell and Turbocharger System

Abstract

This paper identifies and addresses the control challenges associated with simultaneous power and thermal management of a 5-kW-class solid oxide fuel cell and gas turbine combined cycle system. A model predictive controller (MPC) is developed to achieve improved system performance subject to input and state constraints. The subsystem dynamic couplings and control authority limitations under thermal constraints are investigated by both static and dynamic analysis. Through judicious allocation of penalty parameters in the cost function associated with the fuel cell current, anode fuel flow rate, and generator load, several different MPC implementations are derived to explore different control design degree of freedom and their performance is evaluated. Simulation results show the efficacy of the MPC design by demonstrating the fast load transition while maintaining the stack temperature within the allowable operation limits.