Nonlinear observer design for PEM fuel-cell systems using first-order sliding mode techniques

Abstract

This paper presents a nonlinear observer design for Proton Exchange Membrane Fuel-Cell (PEMFC) systems. The aim of the proposed observer is to reconstruct the oxygen excess ratio through the estimation of their relevant states in real time from the measurement of the supply manifold air pressure. A First-Order Sliding Mode (FOSM) differentiation method is adopted to estimate, in finite time, the time derivative of the supply manifold air pressure. By means of the output-state inversion model, the relevant states are reconstructed. The objective of the proposed appproach is to minimize the use of additional sensors in order to reduce the costs and enhance the system accuracy. The performance of the proposed observer is analyzed through simulations considering measurement noise and different stack-current variations. The results show that the nonlinear observer properly estimates in finite time and robustly the oxygen excess ratio.