数据驱动的机器人高精度加工方法
Data-driven High Precision Machining Methods for Robots
孔玲怡 1曹彦生 2戴家隆 3姬煜琦 4王子 4李鹏程3
作者信息
- 1. 南京航空航天大学机电学院,南京 210016;北京新风航天装备有限公司,北京 100854
- 2. 北京新风航天装备有限公司,北京 100854
- 3. 南京航空航天大学机电学院,南京 210016
- 4. 上海飞机制造有限公司,上海 200120
- 折叠
摘要
本文提出了基于数据驱动的机器人高精度加工方法.首先,对实际加工区域进行激光扫描,通过点云特征重构方法实现了加工区域的逆向重构,解决了工件变形后与理论数模不一致的问题;其次,对逆向重构后的加工轨迹进行曲线离散,生成机器人加工点位序列;最后,研究了基于视觉引导的机器人轨迹误差补偿算法,实现了机器人对大型复合材料薄壁件的精确加工.实验结果表明,基于数据驱动的机器人高精度加工方法,加工误差在±0.2mm以内,机器人轨迹误差降低了88.93%,大幅提高了机器人对大型复合材料薄壁件的加工精度.
Abstract
A data-driven robot machining method with high precision was proposed.Firstly,the actual machining area is scanned by laser,and the reverse reconstruction of the machining area was realized by the point cloud feature reconstruction method,which solved the problem of inconsistency between the workpiece and the theoretical numerical model after deformation.Secondly,the curve was discretized after the reverse reconstruction to generate the robot machining point sequence.Finally,the robot trajectory error compensation algorithm based on vision guidance was studied,and the precision machining of large composite thin-walled parts was realized.The experimental results showed that the machining error of the proposed data-driven robot high precision machining method was within±0.2mm,and the robot trajectory error was reduced by 88.93%,which improved the machining accuracy of large composite thin-walled parts.
关键词
大型复合材料薄壁件/三维扫描/逆向重构/视觉伺服/铣边加工/机器人Key words
large composite thin-walled part/three-dimensional scanning/reverse reconstruction/visual servo/edge milling/robot引用本文复制引用
出版年
2024
NETL
NSTL
万方数据