时滞Hopfield神经网络的全局指数稳定性  

Global exponential stability of Hopfield neural networks with delays

作　　者：杨逢建[1] 张超龙[1] 吴东庆[1] 陈新明[1] 

机构地区：[1]仲恺农业技术学院计算科学系,广东广州510225

出　　处：《纯粹数学与应用数学》2008年第1期179-185,共7页Pure and Applied Mathematics

基　　金：广州市科技计划项目(2006j1-C0341)

摘　　要：讨论带有可变时滞的Hopfield神经网络的全局指数稳定性.在非线性激励函数满足Lipschitz条件的假设下,利用推广的Halanay不等式、Dini导数和分析技巧,建立了这类神经网络系统全局指数稳定的几个判别准则.这些判别准则仅仅依赖于系统的参数.This paper is concerned with the global stability of Hopfield neural networks with time-varying delays. Under assumption that the nonlinear stimulate fractions are Lipscliitz Continuous, by means of generalized Halanay inequalities, Dini＇s derivative and functional analysis techniques, several global exponential stability criteria are established, which are only dependent on the parameters of the system.

关 键 词：HOPFIELD神经网络 Halanay不等式 可变时滞 全局指数稳定性 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]