Event-triggered fault detection filter design for uncertain stochastic systems with package dropouts. (November 2019)

• Record Type: Journal Article
• Authors: Ning, Zhaoke; Yu, Jinyong; Wang, Tong
• Is Part Of: Proceedings of the Institution of Mechanical Engineers. Volume 233:Number 10(2019)
• Page Start: 1351
• Page End: 1360
• Publication Date: 2019-11
• Subjects: Fault detection -- event-triggered scheme -- package dropouts -- stochastic systems -- limited network resources; Mechanical engineering -- Periodicals; Automatic control -- Periodicals; Systems engineering -- Periodicals; 621.3
• DOI: 10.1177/0959651819836863
• Languages: English
• ISSNs: 0959-6518

Abstract

In this article, the event-triggered fault detection filter design problem is concerned with uncertain stochastic systems subject to package dropouts. First, a filter structure is constructed to achieve the desired fault detection objective. Second, an integrated model with an event-triggered scheme and a Bernoulli stochastic process are employed to save the limited network resources and describe the package dropouts phenomenon, which always appears in the real network environment. A new sufficient condition is provided to ensure that the obtained residual system is mean square robustly exponentially stable and satisfies the desired detection performance. Then, a novel co-design algorithm is derived to obtain the parameters of filter and event-triggered scheme. Finally, two simulation examples are provided to verify the effectiveness of the proposed design scheme.