John Gaebler, KBR; Juan Gutierrez, KBR; Keric Hill, KBR; Chris Craft, KBR; Paul Billings, KBR; Charles Wetterer, KBR; Jill Bruer, AFRL
https://doi.org/10.64861/ROGR8744
Keywords: Uncertainty propagation, Multi-fidelity, cislunar
Abstract:
Propagating high-fidelity uncertainty distributions for space objects can be a time-consuming process. This effort focuses on efficiently propagating a high-fidelity representation of a satellite probability density function (uncertainty distribution) in cislunar space. An example of an operational use case is the re-acquisition of a cislunar object that has not been observed for a period of a few weeks. Propagating orbital-state uncertainty with stochastic collocation and multi-fidelity (MF) dynamics has shown promise as an efficient method to obtain accurate approximations of large sets of particles or Gaussian mixture model sigma points representing an uncertainty distribution.
Stochastic collocation uncertainty quantification has been used to efficiently predict satellite uncertainty in simple Low Earth Orbit (LEO) and Geosynchronous Earth Orbit (GEO) trajectories with MF dynamics and has been extended to cislunar orbits. The MF process is investigated from the perspective of operational use in the cislunar domain. This robust study of the various tunable parameters establishes guidelines for using MF to propagate uncertainty in cislunar scenarios.
The performance of the MF method is assessed in measurement space on scenarios with realistic a priori uncertainty and long propagation times.
Date of Conference: September 15-18, 2026
Track: Cislunar SDA