Thomas Johnson, Exa Research, LLC
Keywords: Maneuver, Orbit Determination, Characterization
Abstract:
Unknown maneuvers are a common source of degraded orbit determination quality because the maneuver force is not modeled correctly, if it all. We have developed a set of algorithms that perform maneuver detection and immediately start to characterize, reconstruct, and analyze the maneuvers. Multiple hypotheses are carried forward in real time and refined with each new set of tracking data, until the hypotheses have converged or are eliminated. Detection and hypothesis updating is performed using residual and state space analysis using the McReynolds filter-smoother consistency test. Convergence testing uses differential entropy methods. We explain the algorithms used and key considerations for their selection.
The maneuver hypotheses are incorporated into the orbit determination process, resulting in more accurate predicted ephemeris and covariance. Reducing the time delay between maneuver detection and reconstruction helps to maintain track custody. Detailed maneuver characterization allows for pattern of life analysis and more accurate future maneuver modeling.
Simulated and real-world performance results are presented for multiple orbit regimes.
Date of Conference: September 16-19, 2025
Track: Astrodynamics