数学物理学报 ›› 2000, Vol. 20 ›› Issue (3): 400-404.

• 论文 • 上一篇    下一篇

具有可变时滞的Hopfield型随机神经网络的指数稳定性

  

  1. (华中理工大学控制科学与工程系 武汉 430074)
  • 出版日期:2000-05-12 发布日期:2000-05-12

Exponential Stability of Hopfield-Type Stochastic Neural Networks with Variable Delays
 

  1. (Department of Control Science and Engineering, Huazhong University of Science and Technology, Wuhan 430074)
  • Online:2000-05-12 Published:2000-05-12

摘要:

研究了具有可变时滞的Hopfield型随机神经网络的指数稳定性,应用Razumikhin定理与Lyapunov函数,建立了这种神经网络的均方指数稳定与几乎必然指数稳定的两类判据,一类是时滞相关而另一类是时滞无关.

关键词: 神经网络, 指数稳定, Razumikhin定理, Lyapunov函数

Abstract:

paper we investigate the exponential stability of Hopfield-type stochastic neural networks with variable delays, By using the Razumikhin theorems and Lyapunov functions, we present two types of criteria for the exponential stability in the mean square and almost surely exponential stability of Hopfield-type neural networks. One type involves delay dependent results while the other type involves delay independent results.

 

Key words: Neuralnetwork, Exponentialstability, Razumikhintheorem, Lyapunovfunction