Fault Detection for a Class of Discrete‐time Nonlinear Systems Subject to Network‐induced Uncertainties

Xiongbo Wan,Huajing Fang
DOI: https://doi.org/10.1002/asjc.445
IF: 2.4
2011-01-01
Asian Journal of Control
Abstract:This paper addresses the problem of fault detection (FD) for discrete-time systems with global Lipschitz conditions and network-induced uncertainties. By utilizing Bernoulli stochastic variables and a switching signal, a unified measurement model is proposed to describe three kinds of network-induced uncertainties, that is, access constraints, time delays, and packet dropouts. We aim to design a mode-dependent fault detection filter (FDF) such that, for all external disturbances and the above uncertainties, the error between the residual and fault is made as small as possible. The addressed FD problem is then converted into an auxiliary H8 filtering problem for discrete-time stochastic system with multiple time-varying delays. By applying the Lyapunov-Krasovskii approach, a sufficient condition for the existence of the FDF is derived in terms of certain linear matrix inequalities (LMI). When these LMIs are feasible, the explicit expression of the desired FDF can also be characterized. A numerical example is exploited to show the effectiveness of the results obtained. Copyright (c) 2011 John Wiley and Sons Asia Pte Ltd and Chinese Automatic Control Society
What problem does this paper attempt to address?