Adaptive Fault-Tolerant Control for Quadrotor Helicopters’ Pursuit-Evasion Games Under Unknown Opponent Objectives

This paper proposes an adaptive fault-tolerant control (FTC) method for quadrotor helicopters under pursuit-evasion games without the information of evader's objectives. To facilitate the controller design, the quadrotor helicopter dynamics are decoupled into translational and rotational subsystems. For the translational subsystem, an adaptive fault-tolerant game controller is developed via the fault and evader's strategy estimators, enabling the capture of the evader. Moreover, based on the inverse reinforcement learning (IRL), the unknown objectives of the evader are estimated online, achieving the evader's intent recognition. For the rotational subsystem, a fault-tolerant attitude controller is designed by combining the adaptive fault compensation method and reinforcement learning (RL), ensuring the attitude tracking errors converge to a neighborhood of the origin. To improve tracking performance, the proposed attitude controller augments the optimal control law with a robust compensator dependent on the magnitude of the attitude tracking error, which ensures accelerated state convergence and attitude tracking. Finally, the effectiveness of the proposed method is validated through numerical simulations.