基于截断最小二乘和半正定规划的空间非合作目标相对位姿估计

面向空间在轨服务、碎片清除重大应用需求，围绕空间光场复杂、空间目标尺度变化及相对姿态与位置强耦合等难题，提出一种基于截断最小二乘与半正定规划的空间非合作目标相对位姿估计方法。基于TOF相机，首先进行体素下采样，提取快速点特征直方图（Fast Point Feature Histograms, FPFH）特征，考虑目标的局部几何结构尺寸，计算两组FPFH特征的相似度并预测特征之间的对应关系，采用截断最小二乘和半正定规划方法对姿态、位置进行解耦，实现空间非合作目标相对位姿估计。使用非合作目标的3D模型进行仿真验证，仿真结果表明：该方法在三种不同高斯噪声强度下具有较好的估计特性，为后续的工程应用及在轨任务实施提供理论与技术支撑。

Non-cooperative Space Target Relative Pose Estimation via Truncated Least Squares and Semi-definite Programming

Facing the major application requirements of space on-orbit services and debris removal, and focusing on the complex spatial light field, the relative attitude of the target, and the strong coupling between scale, position and attitude, a space non-cooperative target relative pose estimation method is proposed based on truncated least squares and positive semi-definite programming. Through the TOF camera, this paper performs the voxel down sampling, extracts the fast point feature histograms (FPFH) features, considers the local geometric structure of the target, calculates the similarity of two sets of FPFH features, and predicts the correspondence between features using the truncated least squares and semi-positive definite programming method to decouple the position and attitude. And the relative pose estimation of space non-cooperative targets is realized. The 3D model of the non-cooperative target is used for simulation verification The simulation results show that the method has good estimation characteristics under three noise intensities, which provides theoretical and technical support for subsequent engineering applications and task implementation.