Resilient Observer-Based H _∞ Control of Uncertain Nonlinear Networked Control Systems

Abstract

This paper considers the resilient H ∞ dynamic control of nonlinear uncertain systems over a network assuming random communication packet dropouts, a subject which is seldom considered in the current literature for such uncertain systems. The uncertainties in the system and the controller are real, time-varying and norm bounded. Bernoulli distribution with white sequence is used to model the random packet losses with assumed conditions on the probabilitybdistribution. The resilient controllerbdesigned isban observer-basedbdynamic. The resulted closed-loopbsystem is exponentiallybmean squarebstable andbthe H ∞ performance is lessbthan a prescribed levelbγ for allbadmissible uncertainties. Newbsufficient conditionsbfor thebexistence of suchba controllerbare presented and provedbbased onbthe linearbmatrixbinequalities (LMIs)bapproach. A numericalbexample is presentedbto demonstrate and show thebeffectiveness of thebdeveloped theory.