© <2018>. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/

This paper studies a set-based unknown input observer based on zonotopes for discrete-time descriptor systems affected by uncertainties with application to state estimation and robust fault detection. In this paper, two types of uncertainties are considered: (i) disturbances and noise both bounded by zonotopes; (ii) unknown inputs that can be decoupled. In terms of different applications, the observer gain for state estimation is designed to minimize the effects of unknown-but-bounded disturbances and noise as well as state estimation errors. On the other hand, for robust fault detection, in addition to attenuating uncertainties, the designed observer gain is also expected to be sensitive to faults. To achieve this goal, we propose an iterative algorithm to design the fault detection gain. Finally, some illustrative results in an application example show the effectiveness of the proposed algorithms.

Peer Reviewed