An Expectation-Maximization (EM) Algorithm for Orbit Linkage and Determination

• Activity Number: 188 - Contributed Poster Presentations: Section on Physical and Engineering Sciences
• Type: Contributed
• Date/Time: Monday, July 29, 2019 : 10:30 AM to 12:20 PM
• Sponsor: Section on Physical and Engineering Sciences
• Author(s): Jason Bernstein*
• Companies: Lawrence Livermore National Laboratory
• Keywords: mixture model; EM algorithm; maximum likelihood estimation; orbit linkage; orbit determination; time series

Abstract

Satellite orbit determination is important for managing space traffic, avoiding collision, and other reasons. Multiple observations of a satellite are required to determine its orbit, but since multiple satellites can be observed simultaneously, it is not always known which observations are linked to a particular satellite. Hence, orbit determination in this context requires linking observations across time to specific satellites. I discuss a mixture model and EM algorithm for simultaneous orbit linkage and determination. The robustness of this approach to potential challenges, such as an unknown or changing number of satellites over time, limited observations per satellite, and low signal-to-noise ratio, will also be discussed.

LLNL-ABS-765844 This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.