A control method of industrial robot grasping disorderly based on machine vision

In order to improve the grabbing accuracy and stability of industrial robots, a control method of disordered grabbing based on machine vision is proposed. The disordered target is captured by the depth camera, and the hand-eye calibration of the industrial robot is completed. The relationship between the camera and the captured target is transformed by the coordinate transformation matrix, and the position of each disordered individual is defined, which is convenient for the research and control of the dynamics of the manipulator during the subsequent disordered grab process. Then, based on the matching and positioning of the template and the disordered target, the grasping dynamics model of the industrial robot mobile manipulator is constructed. In order to reduce the flexible vibration and hand-eye calibration error, the allowable grabbing area is considered, and the purpose of stable grabbing is achieved by controlling the grabbing area. Finally, the geometric constraint relationship of the contact point under the influence of external force is calculated, and the Hertz contact force model function is used to describe the collision force relationship. By controlling the joint damping, the robot is controlled to grasp the disordered target more stably. Experimental results show that this method can not only meet the requirements of synchronous real-time control, but also effectively improve the accuracy of grasping.