非自治反应扩散Hopfield神经网络鲁棒稳定性  

Global exponential robust stability of Hopfield neural networks with reaction-diffusion trems and S-type distributed delays

作　　者：明清河[1] 杨艳萍[1] 考永贵[1] 

机构地区：[1]枣庄学院数学与信息科学系,山东枣庄277160

出　　处：《云南师范大学学报（自然科学版）》2010年第1期18-22,55,共6页Journal of Yunnan Normal University:Natural Sciences Edition

摘　　要：对具有不确定和S-分布时滞的反应扩散Hopfield神经网络系统,利用拓扑度理论,M-函数,M-矩阵和广义Halanay不等式研究其平衡点的存在性及全局鲁棒指数稳定性,给出易于验证的判别指数稳定性的代数判据.This paper investigates the global exponential robust stability of reaction -diffusion Hopfield neural networks with S - type distributed delays. By means of the topological degree theory M - function, M - matrix, and general Halanay inequality technique, easily verifiable suffcient conditions to guarantee the exponen- tial robust stability of an equilibrium solution associated with temporally uniform external inputs are obtained, without assuming the monotonicity and differentiability of activation functions. The results are new and helpful for designs and applications of neural networks.

关 键 词：S-分布时滞 反应扩散 神经网络 鲁棒稳定性 

分 类 号：O175.12[理学—数学]