Distributed Dynamic Event-Triggered Fault Accommodation for a Class of Nonlinear Uncertain Multiagent Systems

This article presents a distributed dynamic event-triggered fault accommodation leader-following consensus regulation strategy for a class of nonlinear uncertain multiagent systems, remedying time-varying actuator faults and process faults that occur in the local agents and/or are propagated from interconnected neighboring dynamics. The networked system under consideration operates through a bidirectional intercommunication topology with asymmetric weights. Distinct from passive fault-tolerant control schemes and without requiring some a priori knowledge of the specific nature or type of multisource faults, each local agent is equipped with a distributed fault accommodation control law driven by neural network-based learning algorithms that utilize local measurements and information exchanged only with its neighbors. In addition, a less conservative Zeno-free dynamic event-triggered control algorithm with a quantitative compensation mechanism online is developed to achieve reduced data transmission loads, while maintaining satisfactory fault accommodation performance. The closed-loop system stability, minimum interevent time, and cooperative tracking error boundedness in the presence of faults are rigorously investigated and proved. The effectiveness of the proposed scheme is verified through comparative analysis in a simulation example.