具有时滞的双向联想记忆神经网络的稳定性分析

Stability analysis of delayed bi-directional associative memory networks

研究一类具有时滞的变系数双向联想记忆神经网络(BAM)的稳定性问题.由于时滞变系数双向联想记忆神经网络的平衡点不一定存在,分平衡点存在和不存在两种情况进行研究.在BAM神经网络系统存在平衡点时,通过构造Lyapunov函数,结合不等式分析技巧,给出了系统指数稳定的充分条件;在BAM神经网络系统不存在平衡点时,利用数学归纳法给出了系统的吸引域.

The delayed bi-directional associative memory networks with variable coefficients is studied, and a new sufficient condition is presented guaranteeing the exponential stability of the equilibrium when it exists ; In the situation where the system does not have equilibrium, the condition for attractivity is given to work out the attraction set by constructing a Lyapunov function.