State estimation in fuel cell systems : a sliding mode approach

Abstract

In this paper, a nonlinear distributed parameter observer (NDOP) is developed for the estimation of the internal condi-tions of a proton exchange membrane fuel cell (PEMFC). In particular, the estimation of the internal concentration values for different gas species is studied. A novel high order sliding mode (HOSM) estimation technique, based on a distributed parameter model of a PEMFC is developed to reconstruct the internal variables. The observer includes the estimation of the reaction and water transport terms, which are considered as disturbances of the PEMFC estimation model. Since the approach is based on a distributed parameter model representation of the PEMFC, the observer is capable of reconstruct the full internal profile of the estimated variables. The performance and robustness of the proposed observation technique is tested through a simulation example around different operation points.