离散Hopfield神经网络的稳定性分析

Stability analysis of discrete-time Hopfield neural networks

利用能量函数推广了一些关于离散 Hopfield 神经网络稳定性结论,同时给出了稳定点存在的一个充要条件,最后给出了几个有趣的例子,揭示这类网络渐进行为的复杂性。

In this paper, some results for discrete-time Hopfield neural networks are presented, which extend the stability results by the useofenergy function.Further-more,necessary and sufficient conditionfor the existenceand uniqueofnetwork,s equilibrium point is established. Several examples are given to illustrate the complexity of the asymptoticall behavior of this kind of networks.