南京大学学报（自然科学版）2005，Vol.41Issue(2)：215-222,8.
时滞依赖的变延时细胞神经网络的指数稳定性
Delay-dependent Exponential Stability for Cellular Neural Networks with Time-varying Delays
摘要
Abstract
Stability analysis of cellular neural networks (CNNs)has been an important topic in the neuralnetwork field since it has great significance for many applications. The qualitative analysis of the neurodynamics has attracted considerable attention thus far[1～7]. In electronic implementation of neural networks,many problems such as switching delays, integration, and communication delays have arisen. In such a case, a delay parameter must be introduced into the system model. Study of neural dynamics with consideration of delays becomes particularly important in manufacturing high quality microelectronic neural networks. Global stability of delayed cellular neural networks (DCNNs) has been extensively studied[1～11]. Sufficient conditions[5,9,12,13] for global stability of DCNNs have been proposed, but the output of the cell is a piecewise linear function and the time-delay is constant. A wider adaptive range without assuming the output of the cell to be piecewise linear function[10,13] is introduced and the time-delay terms of DCNNs are also constant.Based on the Lyapunov stability theorem as well as some facts about the negative definiteness and inequality of matrices, a new sufficient condition is presented for the existence of a unique equilibrium point and its global exponential stability of the delayed CNNs. This condition imposes constraints on the size of the delay parameter. An illustrative example and its numerical simulation is also given to show the effectiveness of our results.关键词
细胞神经网络/变延时/Lyapunov函数/全局指数稳定Key words
cellular neural networks/time-varying delays/Lyapunov functionals/global exponential stability分类
数理科学引用本文复制引用
梁金玲,黄霞..时滞依赖的变延时细胞神经网络的指数稳定性[J].南京大学学报（自然科学版）,2005,41(2):215-222,8.
