Neural Network Event-Triggered Formation Fault-Tolerant Control for Nonlinear Multiagent Systems With Actuator Faults
Shaocheng Tong, Haodong Zhou, Yongming Li
Abstract
This article deals with an adaptive neural network (NN) formation fault-tolerant control (FTC) issue for nonlinear multiagent systems (MASs) with intermittent actuator faults. Since the controlled MASs contain unknown nonlinear dynamics and unmeasurable states, NNs are applied to model unknown subsystems, and an NN state observer is designed by utilizing intermittent output signals. By the designed state observer and introduced first-order filter technique, a new event-triggered mechanism consisting of both the sensor-to-controller and controller-to-actuator channels is constructed. To avoid the virtual controller nondifferentiability problem by using backstepping control theory directly, this article redesign the virtual controller and controller obtained by the backstepping control technique without considering the event-triggered signals. The developed output-feedback formation FTC scheme can guarantee the controlled MASs are semi-globally uniformly ultimately bounded in presence of the unknown states and actuator faults. Finally, a simulation example confirms the effectiveness of the presented theory and approach.