Event‐triggered consensus secure control for nonlinear MASs under mixed sensor attacks and actuator faults

Summary This article investigates the adaptive neural network (NN) output‐feedback event‐triggered consensus secure control problem for a class of nonlinear multi‐agent systems (MASs) under mixed sensor attacks and actuator faults. Since the considered nonlinear MASs contain unknown nonlinear dynamics, the NNs are first adopted to model unknown agents. Then, a nove NN learning secure state observer is proposed to estimate the sensor attacks and unmeasured states. To reduce unnecessary updating times of the actuator, an event‐triggered mechanism is constructed. By using the backstepping control design technique and the design NN state observer, a NN adaptive output‐feedback event‐triggered consensus secure control scheme is formulated. It is proved that the developed consensus secure control scheme can guarantee the controlled nonlinear MASs are stable and consensus tracking errors converge even under mixed sensor attacks and actuator faults. Simulation and comparative results illustrate the effectiveness of the proposed scheme.