Dynamic threshold computation in fault detection for discrete-time linear systems

Abstract

This paper studies threshold computation in observer-based fault detection for discrete-time linear systems subject to unknown but bounded uncertainties. Based on two different assumptions on uncertainties, we propose two threshold computation methods for residual evaluation. First, we assume that the uncertainties are bounded in known intervals and propose a zonotope-based threshold computation method. Second, the uncertainties are assumed to be norm-bounded and then a threshold computation method using L8 analysis is proposed. Numerical simulations are given to verify and compare the performance of the proposed methods. © 2018 IEEE.