| The live working arm is the development and trend of the electric equipment in the power grid.Compared with the manual live working mode,the live working arm can greatly reduce the working risk factor,improve the automation level and increase the economic benefit,so the live working arm is the hot spot at present.This paper mainly discusses the research of live manipulator vision technology,live manipulator remote operation control technology,these two technologies are the key technology to design and study live manipulator,so the research of this article has important practical significance.In this paper,Firstly,the human factor analysis of the global dynamic vision system of the working manipulator is carried out,and the visual system is analyzed by using the knowledge of human factor engineering,and the relationship between man and machine in the system is designed and improved according to the human characteristics.Secondly,taking UR(Universal-Robots)manipulator as the research object,the additional connecting rod coordinate system and the kinematic model based on D-H(Denavit-Hartenberg)parameters are defined,and the inverse kinematic algorithm of manipulator is established based on geometric method.The effectiveness and accuracy of the kinematic model and inverse kinematics algorithm are verified by experiments.Then,with regard to the calibration of the manipulator vision system,the key content is the calibration of the relative relationship between the camera coordinate system and the manipulator base coordinate system.In this paper,a simple and fast Eye-to-Hand(Eye to Hand)manipulator hand eye calibration method is proposed.This method takes the3D(Three-dimensional)print marking ball fixed at the end of the manipulator as the calibration target.After controlling the manipulator to change the spatial position of the3 D ball many times,the relative position and posture of the manipulator hand and eye can be calibrated.Finally,the calibration experimental results verify the effectiveness and accuracy of the method.The kinematic relationship between VR(Virtual Reality)helmet and UR manipulator is defined and solved.A set of teleoperation global dynamic visual perception positioning system is built in the laboratory.The system includes VR headset,6-degree-of-freedom series robot and binocular depth camera.Based on this system,the master-slave attitude joint tracking,multi-directional observation and positioning of target objects are carried out.The experimental results show that the visual system model can be accurate,and the remote teleoperation control of the binocular camera at the end of the UR manipulator and the multi-directional perceptual positioning of the target object can be realized in real time by the VR headset. |