In this paper, the problems of different finite-time passivity have been concerned for neural network with time-varying delay. Firstly, some different kinds of definitions are given out relating to finite-time passivity of neural networks; then, two different kinds of controllers are established to guarantee neural network finite-time passive, meanwhile, the corresponding conditions are derived by employing inequality techniques and Lyapunov function theories. Moreover, the control protocols for the finite-time stabilization of neural network are presented based on the finite-time passivity theory. At last, some numerical examples are provided to illustrate the effectiveness of resulted conclusions.

本文针对具有时变时滞的神经网络，研究了不同时限被动性的问题。首先，给出了关于神经网络的有限时间无源性的几种不同定义。然后，建立了两种不同的控制器来保证神经网络的有限时间被动性，同时，利用不等式技术和李雅普诺夫函数理论推导了相应的条件。此外，基于有限时间被动性理论，提出了神经网络有限时间稳定的控制协议。最后，通过数值例子说明了结论的有效性。