摘要
讨论一维Hopfield神经网络的局部分岔分析。首先证明了原点在退化情况下为鞍结点,并给出了音叉分岔发生的条件;然后证明了两个非零平衡点在退化情况下是仍然鞍结点,并给出了鞍结分岔发生的条件。
This paper studies the local bifurcation of one-dimensional Hopfield neural networks.Firstly,this paper proves that the origin is saddle-node point in the degenerate condition,and gives the conditionof the pitchfork bifurcation happening.Then it proves that the two non-zero equilibrium points are still saddle nodes in the degenerate condition and gives the condition of saddle-node bifurcation happening.
出处
《成都师范学院学报》
2017年第9期102-105,共4页
Journal of Chengdu Normal University
基金
成都师范学院科研基金(CS16ZB03)
关键词
神经网络
平衡点
鞍结点
分岔
neural networks
equilibria
saddle-node point
bifurcation
