基于双目视觉的非合作目标自主姿态估计方法  

Autonomous Attitude Estimation Method for Non-Cooperative Target Based on Stereo VISION

作　　者：胡启阳[1] 王大轶[2] HU Qiyang;WANG Dayi(Beijing Institute of Control Engineering,Beijing 100190,China;Beijing Institute of Spacecraft System Engineering,Beijing100094,China)

机构地区：[1]北京控制工程研究所,北京100190 [2]北京空间飞行器总体设计部,北京100094

出　　处：《深空探测学报》2019年第4期341-347,共7页Journal Of Deep Space Exploration

基　　金：国家自然科学基金资助项目（61690215,61640304,61573060,61203093）;国家杰出青年科学基金资助项目（61525301）

摘　　要：针对在轨服务任务对于空间非合作目标的自主相对导航需求,提出了一种基于双目视觉的自主姿态估计方法。该方法以Markley变量描述目标的姿态运动,并利用双目相机对目标表面的特征点进行观测,利用特征点的运动规律与目标姿态运动的相关性,并通过容积卡尔曼滤波算法实现了对非合作目标角速度与自旋轴方向的估计。数学仿真验证了所给出的姿态估计方法,并分析了估计精度的影响因素。An estimation method based on stereo vision is proposed for the requirement of autonomous relative navigation to a non-cooperative target in on-orbit servicing mission.First,a dynamic model of Markley variables and a measurement model of stereo vision are built.Second,the relation between the motion of the feature points and the rotation of target is researched.Then the Cubature Kalman filter is used to estimate the spinning angular velocity and direction of the spinning axis.Finally,the validity of the proposed estimation algorithm is verified through numerical simulation.

关 键 词：非合作目标 姿态估计 Markley变量 双目视觉 

分 类 号：V448.2[航空宇航科学与技术—飞行器设计]