摘要
四旋翼无人机的地面移动目标跟踪问题涉及图像序列中的目标跟踪和无人机的飞行控制2个方面内容。针对四旋翼无人机跟踪地面移动目标过程中,出现的目标遮挡和目标从相机视野短暂消失又重新进入视野存在的跟踪失败问题,设计了基于粒子滤波结合P-N在线学习的目标跟踪算法。在此基础上,计算得到目标与无人机间的水平位移作为控制输入,设计了位置-姿态的外内环控制器,在保持无人机飞行稳定性的同时确保目标始终位于机载相机所获得的视野中心区域附近,实现对目标的有效跟踪。仿真和实验结果验证了所设计的四旋翼无人机地面目标跟踪系统的有效性。
Two problems of tracking moving target with quad-rotor UAV are involved, target tracking in image sequences and UAV control. A tracking algorithm based on particle filter combined P-N online learning is designed, since occlusion and target re-entering the camera vision after short disappearance could cause tracking failure. Then, the horizontal displacement between target and UAV is computed as control input, and a position-attitude outer-inner loop controller is designed to ensure the target near the center of camera vision in order to chase the target efficiently. Simulations and experimental results verify the performance of the designed target trackin~ system for quad-rotor UAV.
出处
《控制工程》
CSCD
北大核心
2015年第6期1076-1081,共6页
Control Engineering of China
基金
国家自然科学基金(61304018)