基于低成本MEMS传感器的皮卫星融合定姿算法研究  

A Fusion Algorithm for Pico-Satellite Attitude Determination Using Low Cost Mems-Based Sensors

作　　者：蒙涛[1] 王昊[1] 金仲和[1] 韩柯[1] 

机构地区：[1]浙江大学信息与电子工程学系,杭州310027

出　　处：《宇航学报》2009年第4期1569-1573,1605,共6页Journal of Astronautics

基　　金：教育部新世纪优秀人才支持计划(NCET-06-051)

摘　　要：基于低成本低精度的MEMS陀螺、太阳敏感器、磁强计组合,设计了Unscented Kalman融合滤波算法。同时考虑到可能的故障模式,提出了仅依赖磁强计的备份滤波算法。在姿态参数选取上,采用修正罗德里德参数来描述卫星姿态,降低了滤波器的维数,避免了采用四元数时解算状态误差协方差参数的奇异。利用MEMS陀螺采集的随机噪声数据进行半物理仿真,仿真结果表明:该融合算法能有效对卫星姿态进行实时估计,在个别敏感器失效的情况下,同样能实现较高的精度,满足普通任务的要求。该算法具有工程实现意义,是对低精度MEMS敏感器在低成本皮卫星上应用的有益探索。This paper presents a novel sensing methodology with an unscented Kalman-based fusion algorithm for attitude estimation, using inexpensive MEMs gyroscopes, sun sensor and magnetometer. Considering the potential failure mode, the back-up filter algorithm using only magnetometer is proposed. Three-component modified Rodrigues parameters are used to describe attitude vector, which avoids the singularity of the covariance matrix when using unit quaternion in attitude determination. Using the ground test of gyro drift data, hardware-in-loop simulation shows that the proposed methodology can obtain fine attitude estimation, even only magnetometer is available, the back-up filter can still meets the requirements of the ordinary mission. The work shows profound meaning in the engineering application. This paper can be seen as a positive exploration of low precision MEMS sensors application to low cost pico-satellite.

关 键 词：皮卫星 姿态确定 UKF 修正罗德里格参数 MEMS陀螺 

分 类 号：V448.2[航空宇航科学与技术—飞行器设计]