Spacecraft Custody Maintenance and Maneuver Detection Using Robotic Telescopes and Reachable Sets

Abstract

The increased congestion of the orbital environment necessitates improved ground-based telescope tasking, most forms of which focus on covariance minimization. Limited work has been done on tactical sensor tasking aimed at resolving specific hypotheses of interest, which is the focus of this paper. The discernment of spacecraft maneuvers and maintenance of custody is imperative for improved long-term space domain awareness. As such, a novel telescope tasking algorithm is developed using decision making, evidential reasoning, and reachability techniques. Position reachable set overapproximations are continuously computed assuming that spacecraft are maneuvering using optimal control, which are used to determine whether search or tracking actions are optimal, resulting in improved custody maintenance against worst-case spacecraft maneuvers. Moreover, a new Mahalanobis distance-based binary hypothesis detection method is derived to characterize spacecraft anomaly from a nominal trajectory. A Monte Carlo study is designed to a specified statistical confidence level to examine the accuracy of the algorithm. Lastly, the algorithm is implemented on a robotic telescope in real-time. Thus, using reachable set overapproximations and searches, new evidence to hypothesis belief mappings, a Monte Carlo study, and a real-time demonstration, a novel telescope tasking strategy is presented.