磁定姿近地轨道卫星EKF与UKF算法比较  

Comparison on EKF and UKF forGeomagnetic Attitude Estimation of LEO Satellites

作　　者：杨斌[1] 徐广涵[1,2] 靳瑾[3] 周游[1] 

机构地区：[1]北京航空航天大学电子信息工程学院,北京100191 [2]北京信威通信技术股份有限公司,北京100094 [3]清华大学航天航空学院,北京100084

出　　处：《中国空间科学技术》2012年第6期23-30,38,共9页Chinese Space Science and Technology

摘　　要：某近地轨道(LEO)卫星通过磁强计测量、飞轮控制和磁力矩器卸载,实现三轴姿态稳定。围绕该卫星姿态确定问题,对扩展卡尔曼滤波(EKF)和无迹卡尔曼滤波(UKF)算法进行综合分析对比,最后给出该卫星姿态确定实现方法。分析了地磁模型误差、磁强计测量误差及剩磁的影响,推导了四元数均值及协方差的计算方法。分别引入EKF和UKF进行姿态确定,并讨论了影响算法性能的因素。对两种方法的估计精度、收敛时间及计算需求等指标进行了对比。最后针对该卫星姿态确定要求和硬件条件,选用EKF算法实现姿态确定,满足对地指向精度优于1°,稳定度优于0.1(°)/s的设计指标。A low Earth orbit （LEO） satellite can be stabilized in 3-axis by corporately using a magnetometer, reaction wheels and 3-axis magnetorquer rods. It is challenging to determinate an appropriate attitude for LEO. To address this challenge, the algorithms of extended Kalman filter （EKF） and unscented Kalman filter （UKF） were compared and a solution was proposed by exploring the influence of the geomagnetic model error, the magnetic measurement error and the remanence on the performance of attitude determination. In addition, the calculation of mean and covariance quaternion was derived. The EKF and UKF were introduced to attitude determination, and the factors which have influence on the algorithm performance were discussed. Two algorithms were compared on estimation accuracy, convergence time and computation complexity. With the accuracy requirements and the hardware limitations, the EKF algorithm was finally chosen for the attitude determination. Simulation results show that the earth pointing accuracy of the proposed solution can reach 1° with the stability of 0. 1（°）/s.

关 键 词：扩展卡尔曼滤波 无迹卡尔曼滤波 地磁测量 姿态确定 低轨卫星 

分 类 号：V412.41[航空宇航科学与技术—航空宇航推进理论与工程]