SVD-aided EKF attitude estimation with UD factorized measurement noise covariance

Abstract

This paper describes singular value decomposition (SVD) aided extended Kalman filter (EKF) for nanosatellite's attitude estimation. The development of the filter kinematic/dynamic model, the measurement models of the sun sensors, and the magnetometers used to generate vector measurements are presented. Vector measurements are used in SVD for satellite attitude determination purposes. In the proposed method, EKF inputs come from SVD as the linear measurements of attitude angles and their error covariance. In this step, UD factorizes the attitude angles error covariance, forming the measurements in order to obtain the appropriate inputs for the filtering stage. Results are presented and analyzed in addition to discussion of the sub-step, which is the UD factorization on the measurement covariance. The accuracy of the estimation results of the SVD-aided EKF with and without UD factorization is compared for estimation performance.