Design of an Attitude Estimation Algorithm for a LEO Satellite Based on Multiple Models Adaptive Method and Comparison with EKF

In this paper, a novel and highly accurate attitude estimation method for a LEO satellite is designed. The method is based on multiple model adaptive estimation (MMAE) structure. In this method, the satellite dynamic equation is linearized in a few points in order to increase the computational rate compared with extended Kalman filter (EKF) method. The attitude determination and control system of the satellite is consists of a star sensor, gyroscope and reaction wheels. As known, star sensor is a very power consuming sensor in attitude determination of the satellite; therefore, a lesser power consuming method, using the dynamic model of the satellite along with angular momentum of the reaction wheels, is proposed to estimate the satellite attitude. This method assures the proper operation and the attitude estimation of the satellite in eclipse mode as well. By applying this method, the star sensor is used for a short period of time which reduces power consumption considerably. The performance and effectiveness of the proposed algorithm are investigated through numerical simulations and is compared with extended Kalman filter.