Stability analysis of dynamic multilayer neuro identifier

Author

Yu, Wen

Author_Institution

Departamento de Control Automatico, CINVESTAV-IPN, Mexico City, Mexico

Volume

2

fYear

2002

fDate

10-13 Dec. 2002

Firstpage

1770

Abstract

In the paper, dynamic multilayer neural networks are used for nonlinear system on-line identification. A passivity approach is applied to access several stability properties of the neuro identifier. The conditions for passivity, stability, asymptotic stability and input-to-state stability are established. We conclude that the commonly-used backpropagation algorithm with a modification term which is determined by off-line learning may make the neuro identification algorithm robustly stable with respect to any bounded uncertainty.

Keywords

asymptotic stability; identification; multilayer perceptrons; nonlinear systems; asymptotic stability; backpropagation algorithm; dynamic multilayer neuro identifier; input-to-state stability; nonlinear system; online identification; passivity approach; stability analysis; stability properties; Asymptotic stability; Backpropagation algorithms; Multi-layer neural network; Neural networks; Nonhomogeneous media; Nonlinear dynamical systems; Nonlinear systems; Robustness; Stability analysis; Uncertainty;

fLanguage

English

Publisher

ieee

Conference_Titel

Decision and Control, 2002, Proceedings of the 41st IEEE Conference on

ISSN

0191-2216

Print_ISBN

0-7803-7516-5

Type

conf

DOI

10.1109/CDC.2002.1184779

Filename

1184779