Preliminary Study on Signal-to-Noise Ratio Based Fault Detection and Identification

In the present paper we consider the design of a based fault detection mechanism for a Networked Control System (NCS) feedback loop subject to an SNR constrained. Since the SNR approach is a stationary approach we first discuss the approximation of measuring the channel SNR in a finite time. We then, as a preliminary study, consider the case of a first order plant model with faults such that the plant gain, or the plant unstable pole location, are modified. We observe that the proposed SNR based fault mechanism is capable to first detect and then identify such faults. We then discuss next steps to extend the preliminary results exposed here and conclude with final remarks on the proposed contributions.