Robust, Adaptive and Network Control

An interval observer-based method to diagnose discrete-time systems

A. N. Zhirabok^ab, A. V. Zuev<sup>ba

a</sup> Institute of Marine Technology Problems
^b Far Eastern Federal University, Vladivostok

Abstract: This paper proposes a method for diagnosing linear dynamic systems described by discrete-time models with exogenous disturbances based on interval observers. Formulas are derived to construct an interval observer producing two values of the residual as follows: if zero is between these values, then the system has no faults to be detected by the observer. The case where zero does not belong to the interval between these values is qualified as the occurrence of a fault. The theoretical results are illustrated by an example.

Keywords: linear systems, discrete-time models, interval observers, diagnosis, faults.

Presented by the member of Editorial Board: A. A. Bobtsov

Received: 18.08.2023
Revised: 23.10.2023
Accepted: 01.11.2023

DOI: 10.31857/S0005231023120115

 English version: Automation and Remote Control, 2023, 84:12, 1321–1330