Design and Analysis of Dual-arm Cooperative Robot System for Hole-axis Assembly

In this paper, the robot system of dual-arm cooperative assembly is designed and analyzed in view of the scene of axle-hole assembly in industrial production. The d-h method is used to analyze the kinematics of the 6-dof robot, and the forward and inverse kinematics equations are derived. Then the image is processed by the algorithm to reduce noise and fluctuation interference, so as to obtain more effective information and facilitate the accurate positioning of the visual system. This paper discusses how to identify different color axes by deep learning in halcyon environment. Finally has carried on the preliminary assembly experiment, in the process of assembly, distinguish the color of the parts by using visual sensor characteristics, the main mechanical arm through visual identification, pick one color axis from the arm will be selected according to the main arm shaft select corresponding hole, through the preset trajectory placing it on the desktop, main arm also according to the hole center coordinate point, the position of the moving axis, which are assembled. During the experiment, the coordinate value of the robot arm passing the target position and the visual positioning coordinate value given by the visual system were recorded, and the two were analyzed and compared by Matlab.