关键词: 双向联想记忆, 神经网络, 时变时滞, 鲁棒稳定, 线性矩阵不等式

Abstract: Stability problems are studied for the BAM (bi-directional associative memory) neural networks with time-varying delays, based on the linear matrix inequality technique. Using Lyapunov stability theory and constructing a suitable Lyapunov-Krasovskii functional, two new criteria are given to ensure the global robust stability of the equilibrium point in a BAM neural network with uncertainty. As a result in form of linear matrix inequality, it is easy to verify and independent of the magnitude of the time-varying delays. The magnitude becomes greater if the time-varying delay is short, then the result shows less conservative than the delay-dependent stability as shown in earlier works. A simulation example is given to illustrate the effectiveness of the result.