A control-interval-dependent functional for exponential stabilization of neural networks via intermittent sampled-data control

Abstract In this article, a control-interval-dependent Lyapunov functional is introduced to address the stabilization problem of neural networks under intermittent sampled-data control. By virtue of this Lyapunov functional, the ‘jump’ phenomena of the adjacent Lyapunov functionals at the switching instants can be eliminated without imposing any additional restrictions on the Lyapunov matrices. Combining with the Lyapunov stability theory and inequality estimation techniques, some rigorous analyses on the exponential stability of the resulting closed-loop system are carried out. Then, an explicit expression for controller gain is developed based on the feasibility of certain specified LMIs. Furthermore, a quantitative relationship between the duty cycle of the rest interval and the sampling period is revealed when designing the intermittent sampled-data controller. Lastly, some simulation results are provided to illustrate the effectiveness of the proposed theoretical results.