Target focus capture system based on enhancement learning and impedance variable structure control

Abstract

Abstract For the robotic detecting and grasping of the digestive tract, a cyber-physical system of focus capturing based on a physician's vision and the high-performance robot impedance control method are proposed in this study. The structure of the cyber-physical system for the focus information is given, which includes the physician, the network, the robot, and the digestive tract. The structure is controlled remotely by physicians through a 5G network, and is divided into two layers. The upper layer is mainly for physicians to capture the target of digestive tract focus through enhancement learning, and send instructions to the lower layer robot impedance control set and joint controller through space conversion and 5G network. In this way, the lower layer robot joint can quickly find the target of the digestive tract focus according to the instruction. At the same time, the optimal performance of target capturing, target trajectory processing of Akima, and time-varying nonlinear mapping between robot workspace and joint space are discussed. Finally, the Target Focus Capture system based on enhancement learning and new variable structure impedance control is first proposed for the Doctor-Robot system, and the simulation results show that the remote control method is feasible. In addition, it can be used for reference in COVID-19 prevention and control. At the same time, it is of great help to the examination and operation of unskilled professional physicians.