Design Of Robot Control System Based On Binocular Vision

At present,along with an ageing population and a shortage of labor,"machine for man" has become a major trend in the transformation of the traditional manufacturing industry.Common industrial robots,mainly using manual teaching or offline programming mode,are used in palletizing,welding,painting,assembly and other production and processing occasions that require pre-planned motion trajectories,and their work mode is relatively single,poor operational flexibility,lack of online planning capabilities,cannot be applied to unplanned,random application scenarios,so it is difficult to be widely used to achieve the goal of "machine instead of man".To address these issues,this paper introduces binocular cameras into industrial robots and designs and develops a robot control system based on binocular stereo vision,which provides visual perception capability,online interaction capability,autonomous decision-making capability and real-time operation capability,thus enabling robots to autonomously complete complex production tasks in unplanned scenarios where trajectory paths cannot be planned in advance,further improving the"machine for human" capability,expanding the application scenarios of industrial robots,improving production efficiency and reducing production costs.The thesis first investigates a method for searching and optimizing a robot motion space in which the robot motion is not disturbed by singu,larities,facilitating random online planning of the robot’s trajectory.A human-computer interaction program for the host computer is designed.which can modify the robot motion trajectory online according to the control requirements,or without planning the motion trajectory,and use the EGM interface to control the robot position in real time,thus further improving the robot’s flexibility.The paper investigates robot vision perception based on binocular cameras,and designs a 3D hand-eye calibration program for the robot and the binocular camera,which is used to establish the mapping between the camera coordinate system and the robot coordinate system.Based on the above research basis,the thesis designs and develops a robot control system with autonomous decision-making capability for two engineering application topics.While the application of this system has successfully solved the engineering requirements,it has also fully demonstrated the feasibility of this system in practice.It is hoped that the research in this paper can provide new ideas based on industrial robots for the transformation of traditional manufacturing industries.