The prospect of a permanent sustained presence beyond Earth orbit introduces the problem of accurately identifying, tracking, and managing spacecraft operations, referred to as space domain awareness (SDA), in more distant, dynamically complex regimes. A critical component of SDA is estimating the orbit of a newly detected space object, referred to as initial orbit determination (IOD). Traditional probabilistic approaches to IOD, such as the linear least-squares or information Kalman filters, are limited in representing the uncertainty of IOD estimates, particularly when errors are not normally distributed. In order to define more statistically accurate IOD solutions, this work develops a Gaussian mixture approach that translates a batch of noisy measurements into a probability density function of the object’s dynamic state using Bayes’ rule. The proposed algorithm is demonstrated for two cislunar scenarios, IOD of a cooperative spacecraft using radiometric observations and IOD of a non-cooperative object from an optical observer, and is found to out-perform conventional IOD algorithms.
Gaussian Mixture Batch Estimation for Initial Orbit Determination
J Astronaut Sci
17.06.2025
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Gaussian Mixture Batch Estimation for Initial Orbit Determination
Springer Verlag | 2025
|Probabilistic Initial Orbit Determination Using Gaussian Mixture Models
Online Contents | 2013
|Initial Relative Orbit Determination using Stereoscopic Imaging and Gaussian Mixture Models
British Library Conference Proceedings | 2013
|Gaussian Mixture Approximation of Angles-Only Initial Orbit Determination Likelihood Function
Online Contents | 2017
|