小天体着陆器协方差矩阵分割滤波方法  

Covariance Matrix Segmentation Filtering Method for Small Celestial Body Lander

作　　者：李晋屹 李帅 郭威 刘延杰 李航 邵巍[1,2] Li Jinyi;Li Shuai;Guo Wei;Liu Yanjie;Li Hang;Shao Wei(College of Automation and Electronic Engineering,Qingdao University of Science and Technology,Qingdao 266100,China;Shandong Key Laboratory of Autonomous Landing for Deep Space Exploration,Qingdao 266100,China;Deep Space Exploration Lab,Beijing 100195,China)

机构地区：[1]青岛科技大学自动化与电子工程学院,青岛266100 [2]山东省深空自主着陆技术重点实验室,青岛266100 [3]深空探测实验室,北京100195

出　　处：《航天控制》2023年第6期30-36,共7页Aerospace Control

基　　金：山东省自然科学基金项目(ZR2023MF006);山东省自然科学基金青年项目(ZR2023QF176)。

摘　　要：提出一种小天体着陆器协方差矩阵分割滤波方法,利用惯性测量单元和多个相机估计着陆器运动状态,通过分割协方差矩阵的方式将协方差矩阵隔离更新和传播,并在每次测量更新时使用残差卡方修正方法修正协方差矩阵。通过小天体着陆器着陆仿真测试,与传统扩展卡尔曼滤波算法相比,本文方法在测量噪声估计不准确时具有更好的精度和稳定性。A small celestial body lander covariance matrix partitioning filtering method is proposed in this paper,which utilizes inertial measurement units and multiple cameras to estimate the lander motion state,and isolates the covariance matrix for updating and propagation by partitioning the covariance matrix,and corrects the covariance matrix by using residual cartesian correction method at each measurement update.Through the small celestial body lander’s landing simulation test and comparison with the traditional ex⁃tended Kalman filter algorithm,this method shows better precision and stability when the measurement noise estimation is inaccurate.

关 键 词：小天体探测 自主导航 滤波估计 协方差矩阵分割 

分 类 号：V448.2[航空宇航科学与技术—飞行器设计]