基于3D Zernike矩的巡检器与空间站的相对导航算法

Algorithm for relative navigation between free-flying robots and space station based on 3D Zernike moments

针对航天器相对导航问题,以空间站表面为"特殊地形",提出一种基于大型航天器表面巡检的相对导航算法。首先,运用巡检飞行器上的TOF(Time of Flight)相机测量空间站表面局部点云数据,以该点云数据为实时图,以空间站表面先验点云数据为基准图。然后,利用3D-Zernike矩与三维地形间的一一对应关系,将三维地形匹配转化为基于3D Zernike矩的特征向量匹配。在此基础上求解实时图与匹配上的基准图间的相对位置、相对姿态,从而确定两航天器间的相对导航参数,并通过实验分析了匹配精度及速度的主要影响因素。最后,将该相对导航参数与惯性系统推算的相对导航参数在扩展卡尔曼滤波器的框架下实现信息融合,估计了巡检飞行器与空间站间的相对位置、相对姿态,实验结果表明,相对位置精度优于0.002m,相对姿态精度优于0.1°。

Taking the space station surface as"special terrain",this paper proposes a relative navigation algorithm based on large spacecraft surface inspection.Firstly,the TOF(Time of Flight)camera on the free-flying robot is employed to measure the local point cloud data,regarded as a real-time map,on the surface of the space station.The priori point cloud data on the space station surface is regarded as the reference map.Then the one-to-one correspondence between the 3D Zernike moments and the three-dimensional terrain is used to convert the three-dimensional terrain matching into the feature vector matching based on the 3D Zernike moments.On this basis,the relative position and relative attitude between the real-time map and the reference map are solved to determine the relative navigation parameters between the two spacecraft,and the main factors influencing the matching accuracy and speed are analyzed through experiments.Finally,the relative navigation parameters obtained via terrain matching and the inertial system are fused under the framework of the extended Kalman filter to estimate the relative position and relative attitude between the free-flying robot and the space station.The experimental results show that the estimation accuracy of the relative position is better than 0.002 m,and accuracy of the relative attitude is better than 0.1°.