A synthesis method for the complex-valued associative memory constrained by the attractive domain

Author

Liu, Xiaoyu ; Fang, Kangling ; Liu, Bin

Author_Institution

Eng. Res. Center for Metall. Autom. & Detecting Technol., Wuhan Univ. of Sci. & Technol., Wuhan, China

fYear

2009

fDate

15-18 Dec. 2009

Firstpage

5696

Lastpage

5701

Abstract

A synthesis method for the complex-valued Hopfield network functioned as an associate memory is introduced. Our aim is to guarantee the stored complex-valued patterns are asymptotically stable equilibrium points of the network in the given attractive domain. To this aim we firstly utilize equilibrium equations and Pseudoinverse calculation to obtain the relationship between time constant and connection weights of the network. Then in a given local domain around one equilibrium point we analyze the stability of the complex-valued network and derive the stability condition by Lyapunov theory. The stability conditions are expressed as linear matrix inequalities (LMIs) in the terms of all the varying parameters and hence the network can be easily designed.

Keywords

Hopfield neural nets; Lyapunov methods; asymptotic stability; content-addressable storage; linear matrix inequalities; Lyapunov theory; asymptotically stable equilibrium points; attractive domain; complex valued Hopfleld network; complex valued associate memory constrained; complex valued patterns; equilibrium equations; linear matrix inequalities; pseudoinverse calculation; synthesis method; Associative memory; Control system synthesis; Equations; Hopfield neural networks; Hypercubes; Linear matrix inequalities; Linear systems; Network synthesis; Neural networks; Stability analysis; Complex-valued Hopfield neural network; association memory; attractive domain; linear matrix inequality; stability analysis;

fLanguage

English

Publisher

ieee

Conference_Titel

Decision and Control, 2009 held jointly with the 2009 28th Chinese Control Conference. CDC/CCC 2009. Proceedings of the 48th IEEE Conference on

Conference_Location

Shanghai

ISSN

0191-2216

Print_ISBN

978-1-4244-3871-6

Electronic_ISBN

0191-2216

Type

conf

DOI

10.1109/CDC.2009.5399701

Filename

5399701