Fault-Tolerant Formation Control for Heterogeneous Vehicles Via Reinforcement Learning

This article focuses on the formation control problem of multiple heterogeneous vehicles in air–ground coordination under communication link faults and actuator faults. To address communication link faults among heterogeneous vehicles, a distributed observer is first designed using the air and ground local vehicle information and its convergence is discussed. Using the outputs of the distributed observer and the states of the vehicles, fault-tolerant controller policies are designed for the team of heterogeneous vehicles via reinforcement learning but without knowledge of the vehicle dynamics. Simulation results are included to demonstrate the effectiveness of the proposed fault-tolerant controller.