Interval Parity Relations for Fault Diagnosis in Discrete-Time Stationary Dynamic Systems

The problem of designing interval parity relations to solve the problem of fault detection in stationary systems described by linear and nonlinear dynamic models under external disturbances is studied. The solution is based on a model of the original system that has the minimal dimension, estimates some linear function of the system's output vector, and is insensitive or minimally sensitive to disturbances. The results obtained allow designing interval parity relations based on which the problem of diagnosing faults is solved. The theoretical results are illustrated by an example.