基于AprilTag的分体式飞行汽车自主导引对接方法研究

Research on Auto-Docking Method of Split Flying Vehicle Based on AprilTag Algorithm

摘要:

分体式飞行汽车作为一种新概念空中陆地交通工具，可解决当前频繁发生的城市交通拥堵、城市郊区交通不便捷等问题。针对分体式飞行汽车进行模态转换时涉及的模块间精准导引对接问题提出了一种基于AprilTag的视觉定位导航方案，将AprilTag识别算法的解算结果进行坐标变换后，得到模块间的相对位姿，再结合基于无人机PID控制器的导引对接降落流程设计，解决了因GPS定位误差大而无法达到厘米级精度的导引对接任务需求的问题，并提升了导引对接降落过程的平稳性。最后，在ROS平台利用实物实验验证了该方案的可行性。

Abstract:

As a new concept of aerial and land transportation, split flying vehicles can solve the frequent urban traffic jams and the inconvenient traffic between urban and suburban areas. A vision positioning and navigation scheme based on AprilTag Algorithm is proposed for the precise docking between modules involved in the modal transformation of split flying vehicles. After the coordinate transformation of the result of AprilTag algorithm, the relative position and attitude between modules was obtained. Then combined with the design of the UAV guidance process of landing based on PID controller, which solved the problem that is unable to meet the requirements of the task of guidance docking at centimeter level due to the large GPS positioning error, and the stability of the UAV guidance process of landing is improved. Finally, the feasibility of the scheme is verified by experiments on the ROS platform.