Synchronization of master-slave neural networks with sampled-data control and actuator saturation

摘要

This paper focuses on synchronization of the master-slave neural networks. Under the constraint of sampled-data control and actuator saturation, synchronization criterion of master-slave neural network-based systems has been derived through the dynamic output feedback controller (DOFC). The error signals of master-slave systems are sampled and then transmitted to dynamic output-feedback controller, and an augmented system is modeled as an interval time-varying delay control system with nonlinear items. Through constructing discontinuous Lyapunov function and employing linear matrix inequality approach, sufficient conditions are derived, which guarantee asymptotical stability and accordingly synchronization master-slave systems; on the other hand, under the above synchronization and stability condition, a dynamic output-feedback controller is designed. A numerical example has proved the effectiveness of this method.