[1]蒋秋浩.变时滞神经网络指数稳定性条件[J].南京师范大学学报(自然科学版),2007,30(02):6-10.
 Jiang Qiuhao,Cao Jinde.Global Exponential Stability of Delayed Neural Networks[J].Journal of Nanjing Normal University(Natural Science Edition),2007,30(02):6-10.
点击复制

变时滞神经网络指数稳定性条件()
分享到:

《南京师范大学学报》(自然科学版)[ISSN:1001-4616/CN:32-1239/N]

卷:
第30卷
期数:
2007年02期
页码:
6-10
栏目:
数学
出版日期:
2007-06-30

文章信息/Info

Title:
Global Exponential Stability of Delayed Neural Networks
作者:
蒋秋浩1 2 曹进德2
(1. 中国药科大学数学系,江苏南京210009)
(2. 东南大学数学系,江苏南京210096)
Author(s):
Jiang Qiuhao12Cao Jinde2
1.Department of Mathematics,China Pharmaceutical University,Nanjing 210009,China
2. Department ofMathematics, Southeast University, Nanjing 210096, China
关键词:
细胞神经网络 指数稳定性 变时滞
Keywords:
neural networks exponential stability time-varying delay
分类号:
O175;O177
摘要:
基于时滞细胞神经网络(DCNNs)在图像处理等领域的广泛应用,有关它的研究引起了越来越多学者和专家的关注.早期DCNNs稳定性的结果大多由网络权矩阵的分量构成的代数不等式来表示.运用Lyapunov-Krasovsk ii泛函的方法,研究了DCNNs的指数稳定性,所得充分条件以矩阵的(半)正定形式出现,在实际应用中更加便于验证.与文献中的结果相比较,所得判据适用范围更广.
Abstract:
In recent years, the study of cellular neural networks with delay (DCNNs) draws the attention of more and more specialists due to its extensive app lication in the fields of image p rocessing etc. The ear ly results on stability of DCNNs are usually rep resented by the elements ofweightmatrices of the network. In this paper, by emp loying the Lyapunov-Krasovskiimethod, global exponential stability of DCNNs is in vestigated, and sufficient conditions are obtained in the form ofmatrixwith ( semi-positive) positive defi niteness, and they are easier to be checked in p ractice. Compared to the earlier criteria, our results have more extensive app lications.

参考文献/References:

[ 1 ]  Ensari T, Arik S. Global stability of a class of neural networks with time2varying delay[ J ]. IEEE Transactions on Circuits and Systems -ΙΙ, 2005, 52 (3) : 126-130.
[ 2 ]  周冬明, 曹进德, 张立明. 时滞神经网络全局渐近稳定性条件[ J ]. 应用数学和力学, 2005, 26 (3) : 341-348.
[ 3 ]  Khalil H K. Nonlinear Systems[M ]. New York: McMillan, 1988.
[ 4 ]  L iao X, Chen G, Sancheez E N. Delay dependent exponential stability analysis of delayed neural networks[ J ]. NeuralNet-works, 2002, 15: 855-866.
[ 5 ]  李小平. 算子中立型泛函微分方程稳定性[ J ]. 南京师大学报:自然科学版, 2000, 23 (3) : 20-24.
[ 6 ]  Senan S, Arik S. New results for exponential stability of delayed cellular neural networks[ J ]. IEEE Transactions on Circuits and Systems -ΙΙ, 2005, 52 (3) : 154-158.
[ 7 ]  Chen T. Global exponential stability of delayed Hopfield neural network[ J ]. NeuralNetWorks, 2001, 14 (8) : 977-980.
[ 8 ]  Arik S. An imp roved global stability result for delayed cellular neural networks[ J ]. IEEE Transactions on Circuits SystemsΙ, Fundam. Theory App l, 2002, 49 (8) : 1 211-1 214.

备注/Memo

备注/Memo:
基金项目:国家自然科学基金(6037306) 、江苏省自然科学基金(BK2003053)资助项目.
作者简介:蒋秋浩(1965—) ,硕士,讲师. 主要从事神经网络的教学与研究. E-mail: qiuhaojiang@163. com
通讯联系人:曹进德(1963—) ,教授,博士生导师. 主要从事复杂网络、非线性系统、神经网络的教学与研究. E-mail: jdcao@seu. edu. cn
更新日期/Last Update: 2013-05-05