Precision attitude determination using gyro and star tracker data with a batch least squares estimator

An extended Kalman filter (EKF) has been used for the precision attitude determination (PAD) in the ICESat mission. Some features of the real star tracker data have been sources of accuracy degradation in the PAD. Those systematic error sources require alternative approaches to achieve the accuracy goal. A batch method has been developed using a least squares estimation scheme to address these instrument anomalies. The systematic error sources associated with the star tracker data were effectively reduced or removed in the attitude determination procedure when the batch method was used instead of the EKF.