Pinning stabilization of connected neural networks with event-based couplings

The pinning stabilization problem of connected neural networks (CNNs) is studied in this paper. Event-based sampling protocol is employed for each neural network (NN). An event condition is designed for each neural network to decide the sampling instants. The control signal is also communicated in the event-based fashion. By considering the sampled state as a special time-delay information, a piecewise Lyapunov function is constructed. A stable condition with less conservatism can be derived. Finally, an illustrative example is presented to show the effectiveness of our theoretical results.