Optimal attitude estimation algorithm based on mulitiple star-sensor observations

In most of attitude determinations with methods by using the combination of star-sensors and gyros, the accuracy of attitude determination is high due to the high accuracy of star sensors. However, because of the improper processing of observation model and error, the observation correction ability of attitude filter decreases and the gyro drift error can't be estimated efficiently. Error modeling was proposed based on attitude angle observations, and the optimum installation configurations and observation syncretizing methods were investigated. By making use of the weighted least squares(WLS) preprocessing of observation data, the observation equation changed to be constant. Then, the spacecraft attitude and gyro drift were estimated by using extended Kalman filtering(EKF) with the combination of star-sensors and gyros. The simulation results show the attitude and gyro drift can be estimated accurately by using the presented filter. This method is with less the amount of calculation and of advantage to autonomous on-orbit operation of actual attitude determination system.