具有混合时滞的随机Hopfield神经网络的稳定性分析  

Stability analysis of stochastic Hopfield neural network with mixed time delays

作　　者：吴文娟[1] 刘德友[1] 张静文[1] 刘海涛[1] 

机构地区：[1]燕山大学理学院,河北秦皇岛066004

出　　处：《兰州理工大学学报》2011年第3期89-93,共5页Journal of Lanzhou University of Technology

基　　金：国家自然科学基金(70671088;71071133)

摘　　要：讨论一类具有时变离散时滞和多时变分布时滞的随机Hopfield神经网络的均方鲁棒稳定性问题.假定激活函数既不可微也不严格单调,运用Lyapunov-Krasovskii泛函理论、随机分析理论和MATLAB中的LMI工具箱,推出系统渐近稳定的判别准则.最后通过数值例子论证该结果的有效性.The problem of mean square robust stability of a class of stochastic Hopfield neural network with discrete and multiple distributed time-varying delays was discussed.Assuming there was neither differentiability nor strict monotonicity with the activation function.By using Lyapunov-Krasovskii functional theory,stochastic analysis theory and LMI control toolbox in Matlab,a criterion was derived for discrimination of asymptotic stability of the addressed system.Finally,a numerical example was given to verify the effectiveness of this results.

关 键 词：随机HOPFIELD神经网络 时变离散时滞 多时变分布时滞 有界不确定性 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]