小卫星多传感器融合滤波定姿算法  

A Multi-Sensors Fusion Filtering Algorithm for Small Satellite Attitude Determination

作　　者：段方[1] 刘建业[1] 李荣冰[1] 

机构地区：[1]南京航空航天大学导航研究中心,南京210016

出　　处：《宇航学报》2006年第2期297-300,共4页Journal of Astronautics

摘　　要：虽然扩展卡尔曼滤波被成功应用于许多非线性系统,但由于其对高阶项的截断误差等因素,其用于小卫星多传感器定姿系统时性能受到限制。针对上述问题,本文提出了一种将扩展卡尔曼滤波器与两步滤波器相结合的方法。利用Gibss矢量做姿态参数,避免了四元数在进行迭代时的归一化约束。以太阳敏感器、星敏感器、MEMS陀螺、磁强计为敏感器件,通过仿真,将本文算法与扩展卡尔曼滤波器进行了对比。结果说明,本文算法能将最大稳态定姿误差降低4.4×10-30。Although the EKF has been used successfully in many nonlinear systems, the performance is limited when used in the multi- sensors attitude determination system for small satellite, due to factors such as the truncation of all but first-order terms. A method of combining an EKF and a two-step filter was proposed for this problem. Attitude parameters were Gibbs vectors, so the quaternion norm constraint was avoided while iterating. The sensors were a sun sensor, a star sensor, MEMS gyros and a magnetometer in this paper. The proposed algorithm was compared with an EKF though simulation. Results show that the maximum error in stable stage is decreased by 4.4× 10^-3° for attitude determination and 2.3× 10^-3° for gyro bias.

关 键 词：两步滤波器 卡尔曼滤波 姿态 卫星 

分 类 号：V448.22[航空宇航科学与技术—飞行器设计]