Novel delay-dependent asymptotic stability criteria for neural networks with time-varying delays (Q1019778)

Asymptotic stability for a delay differential model of neural networks is considered. The time-delay \(\tau(t)\) is assumed to be time-varying, nondecreasing, and bounded: \(0\leq\tau(t)\leq h\). For the study of global asymptotic stability of the equilibrium point \(x^*\), the authors construct a Lyapunov-Krasovskii functional. The delay-dependent conditions for global asymptotic stability are obtained in form of LMI. These conditions ensure a large upper bound \(h\) for the time-delay. Two numerical examples are given to indicate the improvement over some existing results.