通有连续时间神经网络的稳定性

The Stability of General Neural Network

研究通有连续时间神经网络的稳定性问题，利用常数交易法、Ｌｉａｐｕｎｏｖ函数法结合不等式分析技巧，获得了通有连续时间神经网络的平衡点是全局ｋ－指数稳定和全局渐近稳定的判别准则。作为特例，也得到了Ｈｏｐｆｉｅｌｄ模型和线性互连模型神经网络的相应稳定性。

This paper,studies the problem of the stability for general neural network.By the method of variation of parameters,the method of the Liapunov function and inequality analysis,the authors established sufficient conditions of the global k-exponential stability and the global asymptotic stability for the neural network.In particular,the stability theorem for Hpofield model and linear exchange link model is obtained.