嫦娥三号着陆器的陀螺在轨标定  

In-flight calibration of the gyros of the Chang'E-3 lunar lander

作　　者：李骥[1,2] 张洪华[1,2] 赵宇[1] 梁俊[1] 张晓文[1,2] 关轶峰[1,2] 黄翔宇[1,2] 

机构地区：[1]北京控制工程研究所,北京100190 [2]空间智能控制技术国家级重点实验室,北京100190

出　　处：《中国科学：技术科学》2014年第6期582-588,共7页Scientia Sinica(Technologica)

基　　金：国家中长期科技发展规划重大专项资助

摘　　要：为了提高月球软着陆过程的导航精度,嫦娥三号着陆器需要在动力下降前进行在轨陀螺安装、刻度因子和常值漂移标定.根据这一特定任务需求,设计了一种新颖的两层滤波算法.第一层滤波算法为惯性姿态估计,用于获得较为准确的星体角速度估值;第二层算法进行陀螺在轨标定,器载计算机使用第一层算法提供的星体角速度估值,利用UD分解扩展卡尔曼滤波,同时估计出陀螺的安装偏差、刻度因子误差和常值漂移.为了提高可观性,着陆器需要按照事先规划的角速度序列进行姿态机动.本文详细介绍了这一方法的原理和设计思路,并进行了公式推导和可观性分析.最后,结合实际飞行数据对该算法的效果进行了说明.In order to improve the navigation precision in the lunar landing process, an in-flight calibration of the gyros should be performed autonomously to estimate those parameters by the Chang＇E-3 lunar lander before the power descending phase began, including misalignment, scale factor errors and biases. For this reason, a new kind of hierarchical filtering algorithm was designed. In the first stage of this algorithm, an attitude estimation method was used to obtain the accurate estimation of the angle velocity of the spacecraft. In the second stage, the onboard computer used a UD Decomposing Kalman filter to estimate the misalignment, scale factor errors and biases of the Gyros simultaneously based on the information from the first filter stage. To increase the system observability, the lander should perform a scheduled attitude maneuver sequence. In this paper, the theory and design philosophy of this method were introduced firstly. Then the procedure of the algorithm and the observability analysis were presented. Finally, the performance of this algorithm was demonstrated using the actual in-flight data.

关 键 词：嫦娥三号着陆器 陀螺标定 安装误差 常值漂移 刻度因子 

分 类 号：V476.3[航空宇航科学与技术—飞行器设计]