航天器非合作目标相对导航的联邦滤波算法研究

Study on Federal Filter of Relative Navigation for Non-cooperative Spacecraft

提出了一种适用于空间机器人在轨对非合作目标进行测量的"激光测距仪+可见光测角相机"组合的联邦卡尔曼滤波相对导航算法。分析了激光测距仪和可见光相机进行非合作目标测量时存在的问题,为此设计了一套包括分别基于测角信息和测距信息的子滤波器,以及进行子滤波器数据融合的主滤波器在内的联邦卡尔曼滤波器,并提出了具体的判据来对子滤波器进行"条件重置"。仿真实验数据表明该联邦卡尔曼滤波器能够在部分目标测量设备的输出出现暂时故障情况下输出较为平稳的相对导航数据,并且滤波算法具有较好的容错性。

This paper presents a federal filter of on-orbit relative navigation for space robot to non-cooperative spacecraft using laser-range-finder（LRF） and vision camera（VC）. It first analyzes the problems when measuring the non-cooperative target using LRF and VC on orbit, then describes the federal filter including main filter for data fusion, and sub-filters for angles and range data filter. Specific criteria in the filter are provided for condition resetting. Simulation experiments result indicates this federal filter can provide smooth relative navigation data and can be more tolerant when some equipment is malfunction and rehabilitation, and system is reeonfigured.