基于视觉的工业机器人离线编程系统的设计

针对传统离线编程系统通用性差、可靠性低和二次开发难度大等问题，开发一套基于机器视觉的工业机器人离线编程系统。基于模块化思想，将该系统划分为机器视觉模块、虚拟环境模块、运动学模块、轨迹规划模块、离线程序模块和外部通信模块。借助机器视觉模块解构视觉系统与机器人末端位姿的坐标映射关系，得到规划机器人运动所需的位姿数据；基于虚拟现实建模语言构建机器人虚拟仿真环境，基于运动学模块与轨迹规划模块将位姿数据转化为机器人的作业〖JP2〗指令；基于离线程序模块与外部通信模块实现控制器指令与虚拟仿真环境的无缝衔接。最后，以一种六轴工业机器人为测试对象验证了该离线编程系统的基本功能。实验结果表明:该系统定位误差最大为0.4 mm，精度高，可满足工业应用需求。

Design of Vision-Based Offline Programming System for Industrial Robots

Aiming at the problems of poor universality,low reliability and difficult secondary development of traditional offline programming system,an industrial robot offline programming system based on machine vision was developed.Based on the idea of modularization,the system was divided into machine vision module,virtual environment module,kinematics module,trajectory planning module,offline program module and external communication module.With the help of machine vision module,the coordinate mapping relationship between vision system and end pose of robot was deconstructed,and the pose data required for robot motion planning were obtained; the robot virtual simulation environment was constructed based on virtual reality modeling language,and the pose data were transformed into robot operation instructions based on kinematics module and trajectory planning module; based on the off-line program module and external communication module,the connection between controller instructions and virtual simulation environment was realized.Finally,taking a six axis industrial robot as the test object,the basic functions of the off-line programming system were verified.The experimental results show that the maximum positioning error of this system is 0.4 mm,which has the characteristics of high precision and can meet the needs of industrial application.