Neural-Network-Based Event-Triggered Sliding Mode Control for Networked Switched Linear Systems With the Unknown Nonlinear Disturbance

The event-triggered sliding mode control (SMC) problem for uncertain networked switched systems with the external unknown nonlinear disturbance is investigated. A neural network (NN) receiving the triggered state is utilized to approximate the external unknown nonlinear disturbance. First, a novel adaptive mode-dependent continuous-time event-triggering scheme (ETS) based on NN weights' estimations is proposed to reduce the burden of the network bandwidth. Then, using the time-varying Lyapunov function method, a novel adaptive NN event-triggered sliding mode controller is established and a dwell-time switching law is obtained, which can guarantee ultimate boundedness, and attain the sliding region around the specified sliding surface for switched systems. Further, a new integral sliding surface that depends on the system states at switching instants and includes the exponential term is proposed. Obtaining the boundary of the sliding mode region relies on the exponential term for continuous-time systems. Moreover, the Zeno behavior can be avoided under the proposed continuous-time ETS by dividing event-triggering signals and switching signals. Finally, a comparative example and a switched Chua's Circuit example are given to illustrate the effectiveness of the proposed method.