摘要
无人机与地面无人车组成空地协同机器人系统,可以显著提高两者作业效率以及机器人系统的自主化程度。针对无人机自主回收过程中存在的定位困难和着陆成功率较低的问题,提出一种适用于空地异构系统自主无人机目标跟踪着陆方法。首先,运用机载视觉算法对降落标志进行目标识别与位置解算,对于无人机难以获得目标的运动状态导致跟踪着陆失败的问题,提出一种基于主动视觉的目标速度估计算法,实现目标无人车运动状态估计。其次,提出分阶段降落策略解决在动态目标上着陆时冗余度较低和目标易短暂丢失的问题,实现无人机对目标的自主跟踪着陆。利用视觉设备辅以无人机仿真实验平台进行实验,结果表明系统各部分都能有效运行,验证了所提出算法和策略的有效性。
Unmanned aerial vehicle(UAV)and ground unmanned vehicles form an air-ground collaborative robot system,which can significantly improve the operational efficiency of the two and the degree of autonomy of the robot system.Aiming at the problems of positioning difficulty and low landing success rate in the process of autonomous UAV recovery,a target tracking and landing method for autonomous UAV suitable for air-ground heterogeneous systems is proposed.Firstly,the airborne vision algorithm is used to detect,identify and solve the position of the landing sign.For the problem that the UAV hardly obtain the motion state of the target,which leads to the failure of tracking and landing,a target speed estimation algorithm based on active vision is proposed to estimate the motion state of the target unmanned vehicle.Secondly,a staged landing strategy is proposed to solve the problems of low redundancy and short-term loss of the target when landing on a dynamic target,and realize the autonomous tracking and landing of the UAV on the target.Experiments are carried out using visual equipment supplemented by the UAV simulation experimental platform.The results show that all parts of the system can operate effectively,which verifies the effectiveness of the proposed algorithm and strategy.
作者
孙凯
于洋
冯迎宾
SUN Kai;YU Yang;FENG Yingbin(Shenyang Ligong University,Shenyang 110159,China)
出处
《沈阳理工大学学报》
CAS
2022年第5期21-28,共8页
Journal of Shenyang Ligong University
基金
辽宁省教育厅科学研究经费项目(LG202009)。
关键词
无人机
目标识别与位置解算
运动状态估计
分阶段降落策略
unmanned aerial vehicle
target recognition and position solving
motion state estimation
staged landing strategy
