Sensor Fault Detection and Isolation via Networked Estimation: Full-Rank Dynamical Systems

This article considers the problem of simultaneous sensor fault detection, isolation, and networked estimation linear full-rank dynamical systems. The proposed is a variant single time-scale protocol based on consensus a-priori estimates measurement innovation. necessary connectivity condition network stabilizing block-diagonal gain matrix derived our previous works. Considering additive faults in presence system noise, error at sensors proper residuals are defined for detection. Unlike many works literature, no simplifying upper-bound noise considered we assume Gaussian system/measurement noise. A probabilistic threshold then detection covariance norm. Finally, graph-theoretic replacement scenario to recover possible loss observability due removal faulty sensor. We examine isolation scheme an illustrative academic example verify results make comparison study with related literature.