Adaptive End-to-End Architecture for Autonomous Spacecraft Navigation and Control During Rendezvous and Proximity Operations

Onboard guidance, navigation, and control (GNC) during rendezvous and proximity operations (RPO) between spacecraft poses unique challenges for associated algorithms. Future missions will require greater on-board autonomy while maintaining in-orbit safety guarantees at varying distances, and scenarios of interest may feature multiple spacecraft which may be cooperative or non-cooperative. This paper presents the conception and development of a new GNC software payload for distributed space systems, which enables safe, autonomous RPO between multiple objects with maximum flexibility and modularity. Navigation algorithms fuse far-range camera images, close-range camera images, differential carrier-phase Global Navigation Satellite System data, and inter-satellite crosslink data to estimate absolute orbits, relative orbits, target pose, and auxiliary states across the entire regime of interest. A control algorithm suite provides optimal maneuver solutions for efficient long-term formation maintenance at far range, centimeter-level rendezvous accuracy at close range, and fast, robust collision avoidance. Prototype simulations featuring cooperative and non-cooperative targets at far, medium and close ranges demonstrate highly capable performance for GNC of distributed space systems, and are a significant step towards achieving complete integration of a flexible autonomous RPO kit for spacecraft.