Design and Implementation of a Dual-Arm Visual Servoing Grasping System for Aerial Manipulation

A new position-based visual servoing (PBVS) scheme for the aerial grasping task of dual-arm aerial manipulators is proposed in this paper. The dual-arm manipulator swing within the x-z plane, so the dynamics model is established in this plane. An improved YOLOv5s-based target detection algorithm is applied using a binocular camera to provide the accurate position and orientation information of the target. Considering the force and torque effects on the hexacopter caused by the displacement of the system center-of-gravity (CoG), a self-balancing motion control strategy is proposed to minimize the impact of the movement of the dual-arm manipulator during the servo grasping process. This ensures the stability of the position and orientation of the hexacopter, so as to adjust the appropriate relative pose between the end effector of the manipulator and the target through PBVS strategy, achieving high-precision servo control of the end effector. In order to perform the grasping task, the active arm moves to the target position, and the passive arm moves to the corresponding position subsequently to grasp the target. The Robot Operating System (ROS) simulation and physical experiment verify the effectiveness of the proposed PBVS strategy.