含时延的联想记忆神经网络的指数吸引域和指数收敛速度(英文)  

Exponential Attraction Domain and Exponential Convergent Rate of Associative Memory Neural Networks with Delays

作　　者：李树勇[1] 徐道义[2] 

机构地区：[1]西安交通大学理学院,西安710049 [2]四川大学数学学院,成都610064

出　　处：《控制理论与应用》2002年第3期442-444,449,共4页Control Theory & Applications

基　　金：supportedbyNationalNaturalScienceFoundationofChina (197710 5 9);EducationBureauofSichuanProvince (0 1LA42 ) .

摘　　要：采用不等式技巧和非负矩阵性质 ,给出了含时延的联想记忆神经网络平衡点的指数吸引域和指数收敛速度估计以及指数稳定的一些判断条件 .Exponential stability of associative memory neural networks with delays, in which there are more equilibria corresponding to different modes, is discussed by employing the inequality techniques and the properties of nonnegative matrices. Some estimation results about the exponential attraction domain and the exponential convergent rate of equilibrium and sufficient conditions for associative memory neural networks with delays to be exponentially stable are obtained. These results can be used not only for evaluation of error_correction capability of associative memory neural networks with delays, but also for the synthesis of such networks.

关 键 词：时延 联想记记忆 神经网络 指数吸引域 指数收敛速度 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]