【摘要】 在作者以前所提出的一种简化现实性神经网络模型一动态神经元网络模型的基础上，进一步考虑了发放神经脉冲引起的兴奋性变化以及每个神经元和其它神经元之间都只有有限联结的性质。使修正后的模型具有更强的生物学真实性。计算机仿真表明由这样的网络产生的类脑电活动对初值极端敏感，具有明显的混沌动力学性质，利用这一点作者们成功地对电脑文档进行了加密和解密，参数和初值的微小改变都使解密失败，从而可望由此得出高度可靠的保密手段。更多还原

【Abstract】 Based on our previously proposed sboltri enlistic neural network model withdynamic neamns, a modal model was developed in this paper, in Which the change of theexchabilitiy of the neamn ac its discharging and the linimed neurons were consider, thus the modal model would be more realistic from a biological view. Computersimulation indicatal that the EEG-like activitieS produced by such model were extrmely sensitive to the initial condition, they had positive Lyapunov exponents and were boUnded in a finite range, thus they had obvious chaotic charactehatics. Taking advantage of there properties,we canmake the files unreadable for susngery and astore them easily. Even very small changeof the parameters or initial values would make the dateding unsucosful, thus, it might beed as a method in communication to keeg the contents secret for those who don’t know thecorrect value of the parametrs and initial values of the model.更多还原