Nonlinear sampled-data observer design via approximate discrete-time models and emulation

Authors:	Arcak Murat, Rensselaer Polytechnic Institute, United States Nesic Dragan, University of Melbourne, Australia
Topic:	2.3 Non-Linear Control Systems
Session:	Nonlinear Observers II
Keywords:	Observers, sampled-data systems, nonlinear systems.

Abstract

We study observer design for sampled-data nonlinear systems usingtwo approaches: (i) the observer is designed via an approximatediscrete-time model of the plant; (ii) the observer is designedbased on the continuous-time plant model and then discretized forsampled-data implementation (emulation). In each case we presentLyapunov conditions under which the observer design guaranteessemiglobal practical convergence for the unknown exact discrete-timemodel. The semiglobal region of attraction is expanded by decreasingthe sampling period. The practical convergence set is shrunk bydecreasing either the sampling period, or a modelling parameterwhich refines the accuracy of the approximate model.