Image-Based Multi-Spacecraft Feature Tracking Control under Visibility Constraints with Observer

• Published in: IEEE transactions on aerospace and electronic systems pp. 1 - 17
• Main Authors: Lin, Zijie, Wu, Baolin
• Format: Journal Article
• Language: English
• Published: IEEE 04.07.2024
• Subjects: Aerodynamics; Aerospace and electronic systems; Aerospace electronics; Cameras; Feature tracking; Image-based visual servoing; Motion constraints; Space vehicles; Spacecraft relative motion control; Target tracking; Three-dimensional displays; Uncooperative object
• ISSN: 0018-9251; 1557-9603
• DOI: 10.1109/TAES.2024.3423848

For multispacecraft missions of tracking noncooperative and tumbling objects, this paper presents an image based control scheme that considers visibility constraints. The primary objective is to achieve six-degree-of-freedom control for steady hovering and relative orientation in two-dimentional image space. Meanwhile, two critical visibility constraints are addressed in this control strategy: 1) ensure that the tracked feature points always remain within the spacecraft's field of view. 2) prevent the tracked features from being occluded by other spacecraft. The former pertains to the motion behavior of a single spacecraft, while the latter deals with the coupled behavior of multispacecraft motion. To accomplish this, the super quadric curve is introduced to describe the boundary of field of view and generate the desired image motion for visibility. A robust tracking controller is designed based on disturbance observer to compensate for dynamic uncertainties. The mutual occlusion model is established in the form of cone, and an active maneuvering strategy is developed for occlusion avoidance .In addition, considering the necessity of features' depth in spacecraft visual servoing, an innovative collaborative depth estimation algorithm is investigated. Though the estimator relies on information from other spacecraft, it does not necessitate multiple spacecraft to simultaneously observe the same area and match the same feature, maintaining the flexibility in the way of monocular vision. The effectiveness of proposed control scheme is validated through comparative simulations of multisatellite observation mission.