数学物理学报

• 论文 • 上一篇    下一篇

随机Hopfield时滞神经网络均方指数稳定性: LMI方法

陈武华; 卢小梅; 李群宏; 关治洪   

  1. 广西大学数学与信息科学学院 南宁 530004
  • 收稿日期:2004-12-14 修回日期:2006-01-04 出版日期:2007-02-25 发布日期:2007-02-25
  • 通讯作者: 陈武华
  • 基金资助:
    广西青年基金(0542032)及国家自然科学基金(10461001, 10462001)资助

Exponential Stability in Mean Square for Stochastic Hopfield Delay Neural Networks: an LMI Approach

Chen Wuhua; Lu Xiaomei; Li Qunhong; Guan Zhihong   

  1. College of Mathematics and Information Science, Guangxi University, Nanning 530004
  • Received:2004-12-14 Revised:2006-01-04 Online:2007-02-25 Published:2007-02-25
  • Contact: Chen Wuhua

摘要: 该文通过系统变换技巧, 构造出新型的Lyapunov泛函. 利用此Lyapunov泛函, 基于线性矩阵不等式, 得到了随机Hopfield时滞神经网络与时滞相关及与时滞无关均方指数稳定性新的充分条件. 数值例子表明, 与已有结果相比, 该文的结果具有较少的保守性.

关键词: 随机神经网络, 时滞, 均方指数稳定性, 线性矩阵不等式

Abstract: By using a technique of model transformation of the system, a new type of Lyapunov functional is introduced. By applying this new Lyapunov functional, a novel delay-dependent sufficient condition of exponential stability in mean square for stochastic Hopfield delay neural networks is derived in terms of
linear matrix inequalities (LMIs). A delay-independent sufficient condition is also presented. Numerical examples show that the proposed method is less conservative than the previous ones.

Key words: Stochastic Hopfield neural networks, Time-delay, Exponential stability in mean square, Linear matrix inequality (LMI)

中图分类号: 

  • 93D