Fault Detection and Isolation of discrete-time Markovian jump linear systems with application to a network of multi-agent systems having imperfect communication channels

• Published in: Automatica (Oxford) Vol. 45; no. 9; pp. 2032 - 2040
• Main Authors: Meskin, N., Khorasani, K.
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.09.2009; Elsevier
• Subjects: Agents (artificial intelligence); Algorithms; Applied sciences; Channels; Communication satellites; Computer science; control theory; systems; Control theory. Systems; Discrete-time Markovian jump systems; Exact sciences and technology; Expert systems; Fault detection; Fault Detection and Isolation (FDI); Fundamental areas of phenomenology (including applications); Geometric FDI approach; Linear systems; Modelling and identification; Network of multi-agent systems; Networks; Physics; Solid dynamics (ballistics, collision, multibody system, stabilization...); Solid mechanics; Network of multi-agent systems; Geometric FDI approach; Discrete-time Markovian jump systems; Fault Detection and Isolation (FDI); Markov process; Markovian system; Probabilistic approach; Transit time; Transmission channel; H infinite control; Solid dynamic; Satellite; Linear time; Modeling; Geometrical model; Multiagent system; Discrete time; Defect; Jump process; Jumping; Formation; Fault diagnostic; Signal to noise ratio
• ISSN: 0005-1098; 1873-2836
• DOI: 10.1016/j.automatica.2009.04.020

This paper deals with the problem of Fault Detection and Isolation (FDI) for discrete-time Markovian Jump Linear Systems (MJLSs). A geometric property related to the unobservable subspace of an MJLS is first presented and the concept of unobservability subspace is introduced. Sufficient conditions for designing an H ∞ -based FDI algorithm for MJLSs subject to input disturbances and measurement noise are presented and developed. Our proposed approach is then applied to the problem of fault detection and isolation in a network of multi-agent systems when imperfect communication channels exist among the agents. A discrete-time communication link with a stochastic packet dropping effect is considered based on the Gilbert–Elliott model and the entire network is modeled as a discrete-time MJLS. Simulation results are presented for formation flight of satellites to demonstrate and verify the effectiveness and performance capabilities of our proposed FDI algorithm.