H∞ filtering for multiple channel systems with varying delays, consecutive packet losses and randomly occurred nonlinearities

• Published in: Signal processing Vol. 105; pp. 109 - 121
• Main Authors: Li, Xiu-Ying, Sun, Shu-Li
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.12.2014; Elsevier
• Subjects: [formula omitted] filter; Applied sciences; Consecutive packet loss; Detection, estimation, filtering, equalization, prediction; Exact sciences and technology; Information, signal and communications theory; Linear matrix inequality; Networked control system; Randomly occurred nonlinearity; Signal and communications theory; Signal, noise; Telecommunications and information theory; Varying delay; Consecutive packet loss; Networked control system; Linear matrix inequality; Varying delay; Randomly occurred nonlinearity; H∞ filter; Performance evaluation; Disturbance rejection; Non linear filter; Data transmission; H; Non linear phenomenon; Networked system; Delay time; H infinite optimization; Packet switching; Filtering; Occurrence frequency; Control system; H infinite control; Independent variable; filter; Sufficient condition; Linear filter; Multiple system; Signal processing; Feasibility; Non linear effect; Transmission loss; Multiple channel
• ISSN: 0165-1684; 1872-7557
• DOI: 10.1016/j.sigpro.2014.05.002

In this paper, we deal with the H∞ filtering problem for a class of multiple channel network-based systems. The system under consideration contains random varying delays, consecutive packet losses, as well as sector-bounded nonlinearities with random occurrence. A group of mutually independent stochastic variables satisfying Bernoulli distributions is introduced to model the addressed system. A linear full-order filter is designed such that, in the presence of all admissible time delays, packet losses and random nonlinearities, the dynamics of the filtering error is guaranteed to be exponentially stable in the mean square sense, and the prescribed H∞ disturbance rejection attenuation level is also achieved. Sufficient conditions are established for the existence of the desired filters. The explicit expression of the desired filter gains can be obtained by solving the feasibility of a linear matrix inequality (LMI). The illustrative examples are given to demonstrate the effectiveness of the proposed method. •We model the measurements with different delay and loss rates in multi-channel NCS.•The randomly occurred nonlinearities are considered as the disturbances.•A linear full-order filter is designed via the LMI approach.•The slack variable is used to reduce the conservativeness in the filter design.