Abstract: Aiming at the accurate recovery of the UAV（Unmanned Aerial Vehicle） on unstable platform especially the swaying platform, a guiding method of optical measurement is used. The key to recovery is to eliminate the impact of swaying platform on measurement, that is to calculate the relative position and attitude between two coordinate systems. For the minimum configuration of three corresponding points, many of the other algorithms are usually singular. It is solved by twice rotation expressed by two Euler angle vectors in this paper. An indoor simulative real-time measurement system is designed based on this principle. Absolute position and attitude of UAV, used for recovery, are calculated through the real-time tracking and measuring and coordinate transformation. A large number of experiments have showed that the method is correct, effective and has a good performance on real time, stability and precision. The impact of swaying platform on measurement and error sources are analyzed in detail, which provides experimental support for the practical application.

Key words: Unmanned Aerial Vehicle（UAV） recovery, optical measurement, unstable platform, sway, real-time, Euler angle

摘要： 针对不稳定平台尤其是晃动平台对无人机的精确回收，采用光学测量的引导方法。回收的关键在于消除平台晃动对测量的影响，即解算坐标系间相对位姿。在三点最小配置情形下，其他多数算法常存在奇异性。利用两组欧拉角向量表示的两次坐标系旋转来解决。基于此原理设计了室内实时测量模拟试验系统，通过实时跟踪、交会测量与坐标转换最终得到无人机在惯性系中的位置和姿态，用于回收。大量试验验证了方法正确有效、实时性和稳定性好、位姿精度高。详细分析了平台晃动对回收的影响及误差来源，为实际应用提供试验支持。

关键词: 无人机回收, 光学测量, 不稳定平台, 晃动, 实时, 欧拉角