节点文献
具有时滞的二阶连续型Hopfield神经网络的周期解
A periodic solution of second continuous Hopfield neural networks with delays
【摘要】 对激励函数为非光滑的二阶Hopfield神经网络的周期解的存在唯一性问题进行讨论.得到了当每个输入函数是 以ω为周期时,该类神经网络存在唯一性ω周期解的一个充分判据.同时,也获得了该类网络的其余各解全局指数收 敛于该ω周期解的一个判据.作为特例,具有时滞的一阶连续型Hopfield神经网络的类似结果也被获得.
【Abstract】 We discuss both the existence and the uniqueness of a periodic solution of second continuous delayed Hopfield neural networks in which all activation functions are nonsmooth. A sufficient criteria is found for both the existence and the uniqueness of ω-periodic solution of the class of the above networks when its each input function is a ω-periodic function. At same time, we obtain also a critera for each solution converges global exponentially to this ω-periodic solution. As a special case, similar results are obtained for the corresponding one-order Hopfield neural networks.
【Key words】 Hopfield neural networks; periodic solution; global exponential stability; Lipschitz condition;
- 【文献出处】 西南民族大学学报(自然科学版) ,Journal of Southwest University for Nationalities (Natural Science Edition) , 编辑部邮箱 ,2004年06期
- 【分类号】TP183
- 【被引频次】5
- 【下载频次】29