Vision based autonomous docking of VTOL UAV using a mobile robot manipulator

摘要

This paper proposes a system that allows the docking of a hovering VTOL UAV (Vertical Take-off and Landing Unmanned Aerial Vehicle) by a mobile manipulator during an exploration or navigation mission for use in heterogeneous robotic team. A single camera configured as eye-in-hand device at the end of the mobile manipulator is used to detect and track the position and orientation of a fiducial marker which is fixed at the bottom of the UAV. By employing visual information in a specified workspace, the position of the UAV is controlled in a predefined 3D coordinate system that allows the UAV to hover in a stationary position with respect to the mobile robot base. At the same time, the end effector is programmed to continuously follow the movements of the UAV in order to prevent the loss of sight. The final contact between the UAV and manipulator is made possible with the approach of the end effector in a controlled manner, while maintaining visual contact with the UAV during the docking process. A special magnetic connection is designed that allows coupling of the two parts that are located at the end effector and at the bottom of the UAV, respectively. The effectiveness of the proposed system is tested through simulations and the results are verified.