摘要
在一类m元离散时滞差分方程神经网络模型中引入了具有明显实际意义的非线性不连续信号传输函数,并利用离散系统的解半环分析这一强有力工具,通过引入一个辅助系统,证明了该模型的每个解或者是最终周期的或者是无界的这一有趣的动力学性质.
This paper proposed a discrete-time difference neural network model for m neurons with delayed feedback, and introduced a class of nonlinear discontinuous signal function. By introducing an auxiliary system and employing semicycle a nalysis as a powerful too1, it is shown that every solution of such a system is either truncated periodic or unbounded.
出处
《经济数学》
2012年第4期32-37,共6页
Journal of Quantitative Economics
关键词
离散神经网络
时滞
最终周期性
周期解
discrete time neural network
delay
truncated periodicity
periodic solution