Conditional Disturbance Compensation-Based Fault-Tolerant Group Consensus Control for MASs: An Event-Based Switching Method

With respect to the group consensus control for multi-agent systems (MASs) with disturbances and actuator nonidentical and unknown direction faults (NUDFs), this paper proposes a novel fault-tolerant control scheme by incorporating an event-based switching function and a saturation function. The event-based switching function is implemented to directly switch to the desired working mode, thereby reducing invalid switching after a reverse fault. Concurrently, the saturation function, acting at the software level, can not only prevent the instantaneous impact of excessive reverse input caused by a reverse fault, but also eliminate input peaking and state chattering after switching the working mode. It is worth mentioning that the proposed switching-based fault-tolerant control method can avoid the control chocks and excessive control gains as in the existing methods by using Nussbaum functions and thus improve the transient performance. Additionally, for disturbances, the designed conditional disturbance compensation mechanism ensures a smaller steady-state error, and less energy consumption by utilizing beneficial disturbances and compensating for harmful ones. Besides, the conditional compensation mechanism also reduces the saturation error. Finally, the effectiveness of the proposed control scheme are verified through a simulation example of uncrewed aerial vehicles (UAVs). Note to Practitioners—Actuator faults, especially reversed faults such as rudder reversals in aircraft or reverse currents in circuits and sensors, can severely disrupt industrial systems. With regard to NUDFs, this paper focuses on the switching function-based fault-tolerant method for high-order high-dimensional MASs, which are applicable to various practical systems such as UAV swarms, multi-robotic manipulators, multi-ship dynamic, multi-manned submersible vehicles, multi-quarter-car suspension, etc. Additionally, the switching function designed in this paper can be easily extended to solve the problems with unknown control directions. On the other hand, the estimator-controller framework provides a solution for achieving group consensus and fault-tolerant control separately, which simplifies the design of the fully distributed controller. Simulations of UAVs confirm the effectiveness of the proposed control approach.