Nonlinear multivariable control of an integrated PEM fuel cell system with a DC-DC boost converter

Abstract

The proton exchange membrane (PEM) fuel cell faces problem when there is an abrupt load demand that is common in transport/vehicle application. This is because of the slow response of underlying electrochemical and thermodynamic processes, abrupt voltage drop proportional to the increased power demand, control system failure, among others. To overcome this shortcoming and to ensure the uninterrupted power delivery, a DC-DC boost converter is coupled with the PEM fuel cell integrated system that additionally includes a few auxiliary subsystems along with a fuel recirculation and a heat exchange unit. For the controlled operation of the entire integrated complex system, a nonlinear observer-based sliding mode controller (SMC) is synthesized. To address a couple of major concerns involved in fuel cell operation, the control objectives are selected at stabilizing the fuel cell power, cathode air flow from the compressor and stack operating temperature. To have the knowledge of unmeasured state information, two nonlinear observers namely, extended Kalman filter (EKF) and sliding mode observer (SMO) are constructed using a low-order process model. The observer-based controller is simulated in MATLAB/SIMULINK environment for their comparative performance. Overall, the closed-loop response of the proposed SMO-SMC and EKF-SMC are found superior to a traditional PI controller.