基于线性矩阵不等式不确定延迟神经网络鲁棒稳定性分析.pdfVIP

30 1 Vol. 30 No. 1 2 0 0 9 1 ACTA ARMAMENTARII Jan. 2009 基于线性矩阵不等式的不确定延迟神经网络 鲁棒稳定性分析 1 2 1 3 林知微 , 仇计清 , 杨洪玖 , 阚军 ( 1, 100081; 2 , 050018; 3 , 100089) : 论文研究了时变时滞的不确定延迟神经网络的鲁棒稳定性问题系统矩阵中包含 数 有界不确定性去掉了时变延迟的导数约束, 这意味着利用 一种快速时变延迟是允许的利用 LyapunovKrasovskii 函数和线性矩阵不等式( LM I) 的方法, 提出一些新的时滞及时滞导数相关的 稳定性准则通过给出一个数值例子, 来阐述文中所提方法的有效性和创新性 : 控制理论; 鲁棒稳定性; 神经网络; 时滞; 不确定系统 : TP273 : A : 2009) 01011 05 LMIBased Robust Stability Analysis for Uncertain Delayed Neural Networks 1 2 1 3 LIN Zhiwei , QIU Jiqing , YANG Hongjiu , KAN Jun ( 1School of Informat ion Science and Technology, Beijing Inst itute of Technology, Beijing 100081, China; 2College of Science, Hebei University of Science and Technology, Shijiazhuang 050018, Hebei, China; 3Chinese Academy of Machinery and Equipment, Beijing 100089, China) Abstract: The problem of robust stability for uncertain neural netw orks w ith timevarying delays was investigated. T he normbounded uncertainties are included in the system matrices. The constraint on timevarying delays is removed, which means that a fast timevarying delay is admissible. Some new stability criteria w hich are correlated w ith time delays and its derivatives w ere presented by using Lya punovKrasovskii function and linear matrix ineguality ( LMI) approaches. A numerical example was given to illustrate the effectiveness and less conservative of the developed techniques. Key words: control theory ; robust st