Nontraditional UKF based nanosatellite attitude estimation with the process and measurement noise covariances adaptation

Abstract

In this study, we add on to our previous researches for non-traditional filtering the investigation of measurement and process noise covariance adaptation and propose an Adaptive Unscented Kalman Filter (AUKF) for nanosatellite attitude estimation. Singular Value Decomposition (SVD) method runs using the magnetometer and sun sensor measurements as the first stage of the algorithm and estimates the attitude of the nanosatellite giving one estimate at a single-frame. Then these estimated attitude terms are given as input to the AUKF. In the result, the attitude and attitude rates of the satellite are estimated reliably in the whole orbital period.