摘要
基于视线测量的航天器相对导航精度会受到相对轨迹形状和滤波算法设计等因素的共同影响。以低轨卫星近距离编队飞行为任务背景,设计了环航飞行、共面漂移和共线保持3种不同轨迹的相对运动模式。对3种模式建立了基于星间非线性相对运动模型的系统状态方程,并引入了J2项地球非球形摄动力的影响;建立了基于视线测量的观测方程,观测量包括星间相对距离、相对俯仰角和相对航向角。结合系统模型和观测模型均为高斯分布的非平稳随机过程的特点,分别在上述3种模式下设计了基于扩展卡尔曼滤波(extended Kalman filter,EKF)和无迹卡尔曼滤波(unscented Kalman filter,UKF)的相对导航滤波算法,对各自的相对运动轨迹进行了数值仿真,并在半物理硬件环境下进行了验证,分析了不同模式下EKF和UKF对于高斯非平稳随机过程的估计精度和稳定性,并结合EKF和UKF的运算复杂度,提出了3种相对运动模式下的滤波器优选方案,对工程设计提供了理论参考。
The precision of relative navigation for spacecrafts based on line-of-sight（LOS） is affected by multiple elements such as the trajectory shape and the filter algorithm. Taking proximity formation flying of low orbit satellites as the mission background, three relative motion patterns called circumnavigation, coplanar drift and collinear holding are designed. The system state equation based on the nonlinear relative motion among the satellites is established, and the J2 perturbation of the earth is also considered; the measurement equation based on LOS is established, and the relative distance, azimuth and elevation among the spaeecrafts are taken as the measurement vectors. Considering that the system model and measurement model both have the feature of Gaussian non-stationary random process, the relative navigation filter algorithms based on extended Kalman filter （EKF） and unscented Kalman filter （UKF） are designed under three relative motion patterns, respectively ; and their relative motion trajectories are numerically simulated and verified in hardware-in-loop simulation environment. The applications of EKF and UKF in Gaussian nonstationary random process are compared in three kinds of relative motion patterns on precision and stability, the calculation complexities of EKF and UKF are considered; and finally the optimal filter scheme under three relative motion patterns is proposed, which can be applied to engineering design.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2012年第6期1201-1209,共9页
Chinese Journal of Scientific Instrument
基金
国家高科技发展计划(2008AA12A221)资助项目
关键词
相对自主导航
扩展卡尔曼滤波
无迹卡尔曼滤波
相对运动轨迹
编队飞行
relative autonomous navigation
extended Kalman filter （EKF）
unscented Kalman filter （UKF）
rela-tive motion trajectory
formation flying
