Title: Reduced-order observer-predictor design for a class of output-delayed systems

Authors: Eric Magarotto; Philippe Dorléans; Jean-François Massieu

Addresses: Department of Physics, Chemistry and Engineering Sciences, UFR of Sciences, University of Caen Normandy, Côte de Nacre, Bd Maréchal Juin, 14-032 Caen Cedex 5, France ' Department of Physics, Chemistry and Engineering Sciences, UFR of Sciences, University of Caen Normandy, Côte de Nacre, Bd Maréchal Juin, 14-032 Caen Cedex 5, France ' University Institute of Technology, University of Caen Normandy, Côte de Nacre, Bd Maréchal Juin, 14-032 Caen Cedex 5, France

Abstract: In this paper, the problem of nonlinear reduced-order observer-predictor design for a class of systems with constant output-delay is discussed. The proposed observer design is an extension of our previous work. This extension, based on a Lyapunov analysis, consists in synthesising a reduced order predictor observer in order to limit calculations as much as possible while guaranteeing an exponential convergence. Comparing to other observer designs, ours is characterised by a simpler structure with only one synthesis parameter. A relation between this parameter and the maximum value of the constant delay, which ensures the global exponential convergence, is established. To show the efficiency of our proposed approach, numerical simulations have been performed on both an academic example and a well-known benchmark.

Keywords: nonlinear systems; observer-predictor; reduced-order; delayed output; Lyapunov stability.

DOI: 10.1504/IJANS.2022.130497

International Journal of Applied Nonlinear Science, 2022 Vol.3 No.4, pp.317 - 336

Received: 16 Feb 2022
Accepted: 05 Sep 2022
Published online: 24 Apr 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article