This paper investigates robust fault isolation and estimation of networked systems, where the plant output is transmitted to the estimator through a network channel with limited communication capacity including signal transmission delay, data packet dropout and measurement quantization. A sufficient condition in terms of affine matrix inequality is developed which guarantees the estimation error below a prescribed bound in the presence of polytopic model uncertainties, time-varying signal delay, networked induced data dropout and quantization. Based on this result, an algorithm for robust fault estimation design is developed by the congruence transformation which transforms the nonlinear matrix inequality to a computable linear matrix inequality (LMI). A numerical example demonstrates that the proposed approach is less constraint restrictive than the existing approaches in literature. Keywords-Networked systems; fault isolation and estimation; linear matrix inequality (LMI); robust quadratic performance; convex optimization I. INTRODUCTION The rapid development of communication technique promotes widespread application of networks. Research on networked control systems (NCSs) has received lots of attention recently, see [1] and the references therein. Many results have been obtained for the stability [2], performance analysis, and controller design of NCSs [3]. However, little work has been done on fault detection and estimation in a network environment.It is known that there are three important problems with the network-based communication channels [4]. The most important one is network -induced delay. In [5], time-delay has been considered as a constant by using the buffer technology. Random time-delay case was considered in [6] where the time-delays were treated as unknown-inputs and thus the problems was solved by designing the observer for unknown inputs. The second interesting aspect of communication channel is data packet dropout (data missing) [7]. This problem was solved in [8] by transferring it to a kind of discrete Markovian jumping process. Another important issue with communication channel is the quantization [9]. In a network environment, signals are usually quantized before being communicated, and the number of quantization levels is closely related to the information flow between the components of the system and thus to the capacity required for transmitting the information.However, to the best of author's knowledge, for fault detection and estimation of networked systems, most references only consider one or two aspects of the communication issues. Few papers simultaneously consider all these three important issues in network-based fault detection and estimation problem. Furthermore, most literature considers discrete system and discrete channel only, and no solution is given for continuous system and discrete channel. The above two critical reasons motivate the current research of this paper.In this paper, the problem of robust fault detection and estimation is investigated for uncertain linear...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.