Stability analysis of discrete-time recurrent multilayer neural networks

Author

Barabanov, Nikita E. ; Prokhorov, Danil V.

Author_Institution

North Dakota State Univ., Fargo, ND, USA

Volume

5

fYear

2004

fDate

14-17 Dec. 2004

Firstpage

4958

Abstract

We address the problem of global Lyapunov stability of discrete-time recurrent multilayer neural networks (RMLNN) in the unforced (unperturbed) setting. It is assumed that network weights are fixed to some values, for example, those attained after training. To apply the method of reduction of attractor estimate, we use the state space extension method to present RMLNN in the form of discrete-time dynamical system. We describe also a new algorithm for checking the global asymptotic stability of RMLNN, which is also based on the method of reduction of attractor estimate, and is much better from the computational viewpoint. An example shows the efficiency of this new algorithm.

Keywords

Lyapunov methods; asymptotic stability; discrete time systems; multilayer perceptrons; neurocontrollers; nonlinear control systems; recurrent neural nets; discrete-time dynamical system; discrete-time recurrent multilayer neural networks; global Lyapunov stability; global asymptotic stability; neural network weights; reduction of attractor estimate; stability analysis; state space extension method; unforced setting; Asymptotic stability; Control systems; Lyapunov method; Multi-layer neural network; Neodymium; Neural networks; Recurrent neural networks; Stability analysis; State estimation; State-space methods;

fLanguage

English

Publisher

ieee

Conference_Titel

Decision and Control, 2004. CDC. 43rd IEEE Conference on

ISSN

0191-2216

Print_ISBN

0-7803-8682-5

Type

conf

DOI

10.1109/CDC.2004.1429592

Filename

1429592