Tatsuyoshi Kurumiya, University of Colorado Boulder, EO Solutions; Cameron Harris, EO Solutions; Connor Segal, EO Solutions
https://doi.org/10.64861/ALYB4524
Keywords: Adaptive Estimation, SDA/SSA, Maneuver Reconstruction, Reachable Sets, Custody Maintenance, Angles-Only Measurements
Abstract:
Maintaining custody of resident space objects (RSOs) under sparse electro-optical observations remains a central challenge in Space Domain Awareness (SDA). In the geosynchronous Earth orbit (GEO) regime, angles-only measurements do not directly observe range, and low-cadence observations can leave maneuver timing and Δv ambiguous. When unmodeled maneuvers occur between observations, standard filtering assumptions become invalid, and multiple post-maneuver trajectories may remain feasible. Maintaining custody of maneuvering objects therefore requires an estimation approach that preserves competing trajectory explanations while supporting maneuver characterization.
This work proposes an adaptive estimation framework for RSOs undergoing bounded impulsive maneuvers under observation cadences representative of existing sensor orchestration systems. The approach uses a sampling-based approximation of the reachable set to guide the post-maneuver search space and constrain hypothesis generation. Adaptive sampling explores maneuver parameters within the feasible region, while Multiple Hypothesis Tracking (MHT) manages competing maneuver explanations across subsequent measurements.
The implemented framework includes a covariance-inflated Unscented Kalman Filter (UKF) baseline, adaptive sampling for hypothesis generation, and an MHT structure with likelihood-based scoring and pruning. Candidate hypotheses are evaluated using measurement consistency, with clustering and local refinement used to identify feasible maneuver modes. Evaluation is performed on representative synthetic impulsive-maneuver scenarios within a high-fidelity SDA simulation environment.
Date of Conference: September 15-18, 2026
Track: Astrodynamics