摘要
用拓扑度和Lyapunov泛函方法,讨论了一类具有时滞的Hopfield神经网络平衡点的存在性及其全局渐近稳定性.所获得的若干判别条件,都去掉了有关文献中关于激活函数的可微性和有界性限制,增强了模型的适用性.
Via the use of topological degree and the Lyapunov functional method, we have discussed the existence of the equilibrium and its globally asymptotic stability to a class of delayed Hopfield neural networks. On this problem, some sufficient conditions are given, in which restrictions for the differentiability and boundness to the activation functions are deleted. This makes the model more applicable.
出处
《生物数学学报》
CSCD
2004年第2期175-179,共5页
Journal of Biomathematics
基金
国家自然科学基金(69974022)资助项目
关键词
神经网络
平衡点
稳定性
拓扑度
Neural networks
Equilibrium
Stability
Topological degree