Robust adaptive observers for nonlinear systems with bounded disturbances

Abstract

Existing adaptive observers for nonlinear systems may generate unbounded parameter estimates in the presence of bounded disturbances. Robust adaptive observers are presented which prevent parameter estimate drift; guarantee the input to state stability property of the error dynamics with respect to disturbances and parameter time-derivatives under persistency of excitation; and achieve asymptotic convergence of state estimation errors for systems in adaptive observer form. Those results are obtained by generalizing a previous result and Barbalat's lemma.