不确定多时滞随机神经网络的鲁棒稳定性  

Robust Stability of Uncertain Stochastic Neural Networks with Multi-delay

作　　者：崔楠[1] 张启敏[1] 

机构地区：[1]宁夏大学数学计算机学院,宁夏银川750021

出　　处：《宁夏大学学报（自然科学版）》2010年第3期216-220,共5页Journal of Ningxia University(Natural Science Edition)

基　　金：教育部重点基金资助项目(208160);宁夏自然科学基金资助项目(NZ0835)

摘　　要：研究了一类不确定随机多时滞神经网络的全局指数鲁棒稳定性.通过引入半鞅收敛定理,运用It微分公式沿系统对构造的Lyapunov泛函进行微分,借助于正定二次型及正定矩阵的相关理论,以线性矩阵不等式的形式给出了具有多时滞的不确定随机神经网络系统几乎必然指数稳定的新的代数判据.The globally exponential robust stability of a class of uncertain stochastic neural networks with multi-delay is investigated in this paper.The problem of stability is studied by taking nonnegative semi-martingale convergence theorem.By means of It differential formula and Lyapunov functions,the stability with multi-delay is developed.Based on positive definite quadratic form and positive definite theory,the new algebraic criterion is presented for almost exponential stability in terms of linear matrix inequalities.

关 键 词：神经网络 半鞅收敛定理 不确定性 

分 类 号：O231.3[理学—运筹学与控制论]