Design of a finite time passivity based adaptive sliding mode control implementing on a spacecraft attitude dynamic simulator. (September 2021)

• Record Type: Journal Article
• Authors: Shahna, Mehdi Heydari; Abedi, Mostafa
• Is Part Of: Control engineering practice. Volume 114(2021)
• Publication Date: 2021-09
• Subjects: Attitude control -- Adaptive -- Sliding mode -- Passivity -- Unwinding -- Virtual velocity -- Fault tolerance -- Uncertainty; Automatic control -- Periodicals; 629.89
• DOI: 10.1016/j.conengprac.2021.104866
• Languages: English
• ISSNs: 0967-0661

Abstract

Abstract: Design of a passivity-based adaptive robust control for attitude tracking of a three-axis satellite is investigated in this paper. By defining a virtual angular velocity for the satellite Kinematics and utilizing the finite time passivity features, it is proved that the developed method drives the system trajectories into the equilibrium point for various signs of the satellite quaternions. Therefore, the closer equilibrium point is always selected and the unwinding problem is resolved. A novel structure is defined for the sliding manifold that uses the selected virtual velocity. Then, a dynamic feedback controller is developed that considers uncertain parameters and faulty actuators (unknown inputs). The upper bounds of unknown inputs and unknown inertia moments are estimated by the developed adaptation mechanisms. A three degrees of freedom dumbbell style dynamic simulator has been developed to provide a rigorous evaluation of the suggested algorithms in a dynamic condition near to space. The proposed algorithms have been implemented for both reaction wheels and thrusters as actuators and the effectiveness of the introduced control methodologies was proved. Highlights: A passivity-based adaptive sliding mode control is investigated. A virtual velocity input is suggested to achieve the finite time passivity. It is ensured that the unwinding problem is resolved. An Attitude Dynamic Simulator is developed to validate the proposed algorithms. Successful performance of the suggested algorithms is proved using experimental results.