Abstract Celestial navigation (CeleNav) has been successfully used during gravity assist (GA) flyby for orbit determination in many deep space missions. Due to spacecraft attitude errors, ephemeris errors, the camera center-finding bias, and the frequency of the images before and after the GA flyby, the statistics of measurement noise cannot be accurately determined, and yet have time-varying characteristics, which may introduce large estimation error and even cause filter divergence. In this paper, an unscented Kalman filter (UKF) with adaptive measurement noise covariance, called ARUKF, is proposed to deal with this problem. ARUKF scales the measurement noise covariance according to the changes in innovation and residual sequences. Simulations demonstrate that ARUKF is robust to the inaccurate initial measurement noise covariance matrix and time-varying measurement noise. The impact factors in the ARUKF are also investigated.
Highlights A novel filter method with adaptive measurement noise covariance is proposed. The impact factors are investigated. The novel filter method simulation on Mars gravity assist flyby is presented. The new method is robust to the inaccurate initial measurement noise covariance matrix and time-varying measurement noise.
An adaptive filter method for spacecraft using gravity assist
Acta Astronautica ; 109 ; 103-111
2015-01-12
9 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
CeleNav , Time-varying , Measurement noise , UKF , Adaptive
An adaptive filter method for spacecraft using gravity assist
Online Contents | 2015
|ENGINEERING NOTES - Spacecraft Planetary Capture Using Gravity-Assist Maneuvers
Online Contents | 2005
|