Data-Driven Fault-Tolerant Control for Attitude Synchronization of Nonlinear Quadrotors

In this article, the data-driven fault-tolerant synchronization control problem is investigated for unknown cooperative quadrotors subject to nonlinearities and multiple actuator faults in the quadrotor dynamics. A distributed observer is provided to estimate the state of the virtual leader. Based on the reinforcement learning theory, the optimal control policy is learned for each quadrotor without any knowledge of the quadrotor dynamic information. Then, the learned control policy is used to construct a data-based fault-tolerant controller to restrain the effects of quadrotor actuator faults. Stability of the constructed controller is proven and the simulation results illustrate the effectiveness of the proposed controller.

Automated summary

This article investigates a data-driven fault-tolerant synchronization control method for unknown cooperative quadrotors, using reinforcement learning and distributed observer. The simulation results demonstrate the effectiveness of this method in controlling the effects of quadrotor actuator faults. The constructed controller's stability is proven in the study.