Reduced neural observers for a class of MIMO discrete-time nonlinear system

Author

Alanis, Alma Y. ; Sanchez, Edgar N. ; Hernandez, Esteban A.

Author_Institution

Dept. de Cienc. Computacionales, Univ. de Guadalajara, Jalisco, Mexico

fYear

2009

fDate

10-13 Jan. 2009

Firstpage

1

Lastpage

6

Abstract

A nonlinear discrete-time reduced order neural observer for the state estimation of a discrete-time unknown nonlinear system, in presence of external and internal uncertainties is presented. The observer is based on a discrete-time recurrent high order neural network (RHONN) trained with an extended Kalman filter (EKF)-based algorithm. This observer estimates the state of the unknown discrete-time nonlinear system, using a parallel configuration. The paper also includes the stability proof on the basis of the Lyapunov approach. To illustrate the applicability simulation results are included.

Keywords

Kalman filters; Lyapunov methods; MIMO systems; discrete time systems; learning (artificial intelligence); nonlinear control systems; observers; recurrent neural nets; reduced order systems; uncertain systems; Lyapunov approach; MIMO system; discrete-time system; extended Kalman filter; nonlinear system; parallel configuration; recurrent high order neural network; reduced order neural observers; stability proof; state estimation; Control systems; Kalman filters; MIMO; Mathematical model; Neural networks; Nonlinear systems; Observers; Recurrent neural networks; State estimation; Uncertainty; Discrete-time nonlinear systems; HIV model; Kalman filtering learning; Recurrent high order neural networks; Reduced order neural observers;

fLanguage

English

Publisher

ieee

Conference_Titel

Electrical Engineering, Computing Science and Automatic Control,CCE,2009 6th International Conference on

Conference_Location

Toluca

Print_ISBN

978-1-4244-4688-9

Electronic_ISBN

978-1-4244-4689-6

Type

conf

DOI

10.1109/ICEEE.2009.5393313

Filename

5393313