Fault detection and estimation in systems with scarce measurement

In this paper, the fault detection and estimation in discrete-time systems whose outputs are measured scarcely and irregularly with different sensors is addressed. An observer that estimates the states and the faults is used in which the updating gain depends on the sensors availability. The process is assumed to be affected by disturbances of known bounds, and the proposed approach allows to achieve a compromise between the time response of the fault detector and the disturbances attenuation (related to the faults detection threshold). Two strategies are proposed: in the first one, the threshold is fixed, and the response time is minimized. In the second one, the response time is fixed, and the threshold is minimized. In both cases, the observer design is based on H*infin; norms minimization via LMI.