Flight Demonstration of Autonomous Noncooperative Rendezvous in Low Earth Orbit

Abstract

This paper presents ultimate design, implementation, and in-flight performance of the spaceborne guidance navigation and control system which enabled the Autonomous Vision Approach Navigation and Target Identification (AVANTI) experiment; a flight demonstration developed by the German Space Operations Center (GSOC) of the German Aerospace Center (DLR) and carried out in November 2016. Designed to prove the viability to perform far-to mid-range proximity operations with respect to a noncooperative flying object using only optical angle measurements, AVANTI realized the first autonomous vision-based rendezvous to a passive target spacecraft in low Earth orbit. Within this experiment, the DLR Earth-observation BIROS satellite approached down to less than 50 m of inter-satellite distance the BEESAT-4 CubeSat, previously released in orbit by BIROS itself. To this end, a dedicated spaceborne formation-flying system carried out relative navigation and maneuver planning tasks. Moreover, it took over BIROS orientation and maneuvering capabilities to steer the spacecraft along a passively safe rendezvous trajectory. During AVANTI, the images taken by BIROS constituted the only source of relative navigation information. In the absence of external, independent, and precise navigation data of the target satellite, AVANTI performances have been assessed against the ground-based post-facto reprocessing of the images collected in flight.

Kinematic Approximation of Position Accuracy Achieved Using Optical Observations of Distant Asteroids

“…These ambitious objectives are driving significant investment in autonomous navigation technologies that ensure accurate terminal delivery of the spacecraft to its target. Major autonomous navigation efforts are currently underway to support asteroid and comet sample retrieval [8], soft landing on planets and large moons [9][10][11][12], hazardous asteroid deflection [13], formation flying [14], and onorbit rendezvous between spacecraft [15][16][17].…”

In-Plane Formation Reconfiguration with Radial Maneuvers

“…I MPULSIVE control for formation reconfiguration has been continuously studied in the past decades because it is the key to substantializing a space mission with multiple satellites. The methods can be roughly divided into two parts: one with relative Cartesian coordinates [1][2][3] and the other with orbit elements difference [4][5][6][7][8][9][10][11][12]. Concentrating on the latter part, Schaub and Alfriend [6] proposed an analytical solution consisting of three impulses.…”

Adaptive Extended Kalman Filtering Strategies for Autonomous Relative Navigation of Formation Flying Spacecraft

“…Instinctively the risk of collision should preclude the need for two high-velocity objects to operate in such close proximity. However, while concerns of satellite collisions and interference were indeed warranted during the early years of the space age, recent demonstrations of advanced Guidance, Navigation and Control (GNC) systems onboard the PRISMA [1], CanX-4 and CanX-5 [2], and AVANTI [3,4] spacecraft maintained control accuracy at the sub-meter level, presenting a new host of possibilities for the safe utilization of co-orbiting, cooperative formations. The beneĄts of formation Ćying stem from the distribution of payload systems and operational functionality across multiple, smaller spacecraft.…”

“…Precise formation keeping control was required during science-collection operations, which required knowledge of the spacecraft separation to within 100 m. This accuracy was accomplished by onboard navigation software that used GPS measurements, an EKF, and a high Ądelity onboard dynamics model [32]. Thereafter, the Autonomous Vision Approach Navigation and Target IdentiĄcation (AVANTI) experiment [3] was performed in June 2016, using the Bispectral InfraRed Optical System (BIROS) and Berlin Experimental and Educational Satellite 4 (BEESAT-4) spacecraft of the DRL Firebird mission [33,34]. These spacecraft used line-of-sight measurements and an EKF to estimate the relative spacecraft state in terms of Relative Orbital Elements (ROEs), but were only able to achieve relative navigation accuracy at the meter level due to observability issues with using line-of-sight methods [35].…”

“…The system can further be simpliĄed if we assume r t ≫ ¶x, y, z♢, such that we can approximate r c with a Taylor Series expansion. Taking only the Ąrst-order terms for the expansion of r c , neglecting higher-order terms and knowing n = µ/r 3 t , we get…”