ASILO-Based Active Fault-Tolerant Control of Spacecraft Attitude with Resilient Prescribed Performance

In this study, an active fault-tolerant control problem was addressed for a rigid spacecraft in the presence of unknown actuator faults, uncertainties, and disturbances. First, an adaptive sliding mode iterative learning-based observer (ASILO) is proposed for diagnosing and reconstructing unknown faults. It achieves greater accuracy and rapidity while consuming less computing resources by constructing adaptive gain based on an auxiliary error. Specifically, it significantly improved the computational efficiency by 76% compared with the Strong Tracking Kalman Filter while achieving a similar accuracy. It also enhanced the accuracies relative to the traditional ILO and adaptive ILO by 67% and 36%, respectively, and demonstrated 82% and 52% increases in rapidity. Then, fault-tolerant control with resilient prescribed performance (RPP) that can adapt to changing initial conditions and adaptively adjust performance constraints online by sensing faults and error trends is proposed. It avoided the control singularity by constructing adaptive resilient boundaries with almost no impact on the computational overhead. It significantly improved the performance and conservatism. Finally, the robustness and effectiveness of the proposed strategy were demonstrated by numerical simulations.