Modular Vision-Based Navigation System for Autonomous Rendezvous and Docking in Geostationary Transfer Orbits
Year: 2025
Type: conference-paper
Abstract: Nowadays, the growing demand for autonomous in orbit servicing (IOS) missions, such as inspection, refueling, and docking, has made vision based autonomous navigation a critical element of Guidance, Navigation and Control (GNC) systems.In particular, rendezvous operations require robust solutions across a wide range of conditions, from the initial detection to the final docking phase.This work presents a modular vision-based navigation system with a dual-mode architecture tailored for fully autonomous rendezvous and docking.Specifically, in the far range phase, a dynamic mode switching logic evaluates the alignment between the target geometric center and its center of mass.When alignment is detected, Line of Sight (LoS) only measurements are employed.Otherwise, a pixel based range estimation is added.This adaptive measurement strategy feeds an Extended Kalman Filter (EKF), ensuring robust relative state estimation across varying distances and visual conditions.Within 15 meters, the system transitions to close range navigation using ArUco markers on the target docking face.Marker detections enable 6-DoF pose estimation via a Perspective-3-Point (P3P) algorithm, refined through the Unscented Quaternion Estimator (USQUE).Moreover, a camera switching mechanism ensures continuous marker visibility during final approach.Validated within the GEORyder project, this system offers adaptive far range estimation, high precision close range tracking without external aids, and a scalable architecture for future IOS applications.
Institution: Politecnico di Milano
Language: English
Subfield: Aerospace Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7133584929