Observation and Maneuver Detection for Cislunar Vehicles

Abstract

International interest in the sustained development of cislunar space will generate traffic and debris in the region which requires monitoring; similar to how current space situation awareness is necessary for the traffic and debris near Earth. There are many challenges associated with developing a cislunar situation awareness program, but 2 primary issues addressed by this paper are observational strategies and maneuver detection methods. This work proposes an observational strategy that utilizes the ballistic Optimal Control Based Estimator (OCBE) to filter measurements from cislunar optical observers. To reduce numerical issues associated with filtering, new modifications to the ballistic Optimal Control Based Estimator (OCBE) are introduced that preserve the OCBE update equations in Square Root Information (SRI) space. This new derivation produces a more stable version of the ballistic OCBE which is beneficial for filtering larger data sets with non-linear motion and measurements. Applying the SRI OCBE to the estimation problem it was found that only a single L2 observer with angle and angle-rate measurements provided sufficient information for consistent estimation. Then a newly developed maneuver detection method is presented to statistically identify maneuvers. The method applies a binary hypothesis test to the optimal control policy of the ballistic OCBE to quantify mismodeling. This method was tested given a impulsive maneuver policy with a mean of 50 mm/s and standard deviation of 15 mm/s, and 194 out of 200 tests correctly identified if a maneuver occurred. The OCBE control policy also provided appropriate impulse estimates of mismodeling, which may be used to reconstruct maneuvers in future work. Together, the proposed observation and maneuver detection methodology yields reliable tracking and provides a statistical framework to detect maneuvers.