随机时滞神经网络的指数稳定性(英)  

Exponential Stability of Stochastic Delay Neural Networks

作　　者：付莉红[1] 万钧力[1] 沈轶[1] 

机构地区：[1]湖北三峡学院理工学院机电工程系,宜昌443003

出　　处：《三峡大学学报（人文社会科学版）》1999年第5期15-18,共4页Journal of China Three Gorges University(Humanities & Social Sciences)

基　　金：国家自然科学基金

摘　　要：　本文研究了具有不确定性随机时滞神经网络的指数稳定性，利用Ｌｙａｐｕｎｏｖ 泛函鞅不等式和Ｂｏｒｅｌ－ Ｃａｎｔｅｌｌｉ 引理，建立了均方指数稳定与几乎必然指数稳定的充分判据，这些判据是时滞无关的且仅要求检验某一不等式－ 这些结果是最近某些文献中结果的推广－ 最后。In this paper,we investigate exponential stability of stochastic delay neural network with uncertainty. By utilizing the Lyapunov functional, martigale inequality and Borel-Cantelli lemma,we establish sufficient criteria for the exponentially stable in the mean square and almost surely exponentially stable. These criteria are independent of delay and only require verification of a certain inequality .Our results are generalizations of some recent ones reported in the literature. In final, a numerical example is given to illustrate the effectiveness of the obtained criteria.

关 键 词：随机时滞神经网络 指数稳定性 BOREL-CANTELLI引理 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]