本文研究了一类具偏李普希兹连续和单调增加激活函数的神经网络绝对指数稳定性问题 .所得结果归结为如果联接矩阵T的负矩阵是一个非负对角元的H矩阵 ,那么在任意输入向量和网络参数的条件下 ,所选激活函数只要是偏李普希兹连续且单调增加的 ,广义动态神经网络绝对指数稳定 .该广义动态神经网络包含常用的Hopfield神经网络 ,双向联想记忆神经网络和细胞神经网络作为其特殊情形 ,所得结论推广了现有文献中的有关结论 In this paper, we study the absolute exponential stability of a class of NN with monotone increasing and Lipschitz continuous increasing. The results are summarized as follows: if the negative matrix of the connection matrix T is a non-negative diagonal H matrix, then at any input Vector and network parameters, as long as the selected activation function is Lipsitz continuous and monotonically increasing, the generalized dynamic neural network is exponentially stable.The generalized dynamic neural network includes the commonly used Hopfield neural network, two-way associative memory neural network and cells As a special case of neural network, the conclusion draws the conclusion in the existing literature