Authors: Gary Goff; Jonathan T. Black; Joseph A. Beck
Addresses: Department of Aeronautics and Astronautics, Air Force Institute of Technology, Dayton, OH, USA ' Department of Aerospace and Ocean Engineering, Virginia Tech, Blacksburg, VA, USA ' Air Force Research Laboratory, Wright-Patterson Air Force Base, Dayton, OH, USA
Abstract: Given limited observation data, an initial orbit fit is often inaccurate with a large associated error covariance. The affect of the quantity of observations on the error covariance and orbit fit is explored. This paper investigates the entire initial orbit determination estimation chain using simulated ground radar data and filter estimation techniques. A probabilistic approach is developed to determine a specific number of radar observations to estimate the future location of a non-cooperative object within an ellipsoidal error tolerance. Specifically, the article evaluates the estimation chain from observations to orbit fit for covariance realism and develops a numerical sampling technique to calculate the probability of meeting a covariance requirement given a set number of observations, orbit type, and radar performance.
Keywords: orbit determination; covariance estimation; non-cooperative tracking; short-arc initial orbit; collision avoidance; covariance realism; short arc radar observation; covariance intersection; error covariance; orbit fit; simulation; ground radar; filter estimation; ellipsoidal error tolerance; obstacle avoidance.
International Journal of Space Science and Engineering, 2015 Vol.3 No.1, pp.50 - 67
Received: 20 Dec 2014
Accepted: 04 Feb 2015
Published online: 12 May 2015 *