Fault detection for discrete time-delay networked systems with round-robin protocol in finite-frequency domain

The paper considers the fault detection problem over finite-frequency domain for a class of discrete time-delayed networked systems subject to round-robin protocol. This protocol is a periodic one to schedule the information transmitted via a shared communication network with an intent to prevent th...

Full description

Saved in:
Bibliographic Details
Published inInternational journal of systems science Vol. 50; no. 13; pp. 2497 - 2509
Main Authors Ju, Yamei, Wei, Guoliang, Ding, Derui, Zhang, Sunjie
Format Journal Article
LanguageEnglish
Published London Taylor & Francis 03.10.2019
Taylor & Francis Ltd
Subjects
Online AccessGet full text
ISSN0020-7721
1464-5319
DOI10.1080/00207721.2019.1671530

Cover

More Information
Summary:The paper considers the fault detection problem over finite-frequency domain for a class of discrete time-delayed networked systems subject to round-robin protocol. This protocol is a periodic one to schedule the information transmitted via a shared communication network with an intent to prevent the data from collisions. In this scenario, the error dynamics can be regarded as a periodic protocol-induced system. The aim of the problem addressed is to design a fault detection filter that (1) the sensitivity of the residual to the fault is enhanced by means of a maximised index and (2) the effect from the exogenous disturbances on the residual is decreased with considering a minimised index. By resorting to a Lyapunov function, an improved version of generalised Kalman-Yakubovich-Popov lemma (GKYP) is first developed for protocol-induced periodic system. Then, in light of such a lemma, sufficient conditions are derived to guarantee both the fault sensitivity and the disturbance attenuation capability by solving a set of matrix inequalities. Finally, the usefulness of the presented fault detection algorithm is utilised via a numerical simulation.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0020-7721
1464-5319
DOI:10.1080/00207721.2019.1671530