基于视觉技术的铸件打磨机器人设计

Design of Casting Grinding Robot Based on Vision Technology

通过分析铸件毛坯现有打磨技术的劣势,提出自动扫描三维重构铸件模型,获取工件3D点云数据,然后根据打磨工艺自动规划路径并进行离线编程,获得机器人运动程序,最后按照工艺路径自动完成铸件打磨的工艺方案。在综合考虑成本、功能、力学及干涉等各方面因素的基础上,确定打磨装备的整体布局形式,并对专用恒力打磨机构、视觉系统等关键部件作了详细设计与阐述。利用现有的机械设备、工业相机及线激光组成的视觉系统对实验工件进行三维模型重构,结果显示:模型尺寸精度在±0.1 mm以内,扫描速度为1 000 mm/min。打磨实验显示:打磨后工件表面粗糙度值小于Ra6.3μm,整个打磨表面平整度也满足检测需要,单台设备打磨效率达到人工的10倍左右。

By analyzing the disadvantages of the existing grinding technology of casting blank, a new process plan was put forward. In this plan, the 3 D reconstruction casting model was scanned automatically to obtain the 3 D point cloud data of the workpiece, and then the path was automatically planned according to the grinding process and offline programming was carried out to obtain the robot motion program. Finally, the casting grinding was completed automatically according to the process path. Based on the comprehensive consideration of cost, function, mechanics and interference, the overall layout of the grinding equipment was determined, and the special constant force grinding mechanism, vision system and other key components were designed and described in detail. Using existing mechanical equipment, industrial camera and line laser vision system to reconstruct the three-dimensional model of the experimental workpiece, the results show that the size accuracy of the model is within ±0.1 mm, and the scanning speed is 1 000 mm/min. The grinding experiment shows that the surface roughness of the workpiece after grinding is less than Ra 6.3 μm, and the whole grinding surface flatness also meets the detection needs. The grinding efficiency of single equipment is about 10 times of that of manual.