期刊文献+

基于点云和六维力信息的机器人螺栓装配方法

Robot bolt assembly method basedon point cloud and six-dimensional force information
在线阅读 下载PDF
导出
摘要 针对目前位姿不确定场景下的螺栓装配存在工人劳动强度大、装配结果一致性差和装配效率低等问题,面向钢厂轧钢机换辊过程中拆装螺栓连接的工业场景,融合工业机器人、立体视觉定位和力控制技术,提出了一种基于点云和六维力信息的机器人螺栓装配方法。通过一种视觉引导的视觉定位策略和一种基于众数的自适应直通滤波算法,动态调整滤波阈值消除环境噪声干扰,获取目标孔和螺栓头的位姿信息,在此基础上进行基于力信息的螺栓装配,提出一种“米”字形变步长搜索策略,实现螺栓端面被目标孔包含;采用一种导纳控制策略,实现了螺栓顺应性插入目标孔;提出一种基于垂径定理的孔心定位方法,实现了螺栓轴线和目标孔轴线的重合。该方法突破了传统示教编程对结构化环境的依赖,为复杂工业场景下的智能装配提供了可扩展的技术框架,结果表明,提出的方法在未知场景下进行螺栓装配的成功率大于90%,耗时低于5 s,该方法有效实现未知环境下的螺栓装配,为螺栓装配自动化提供理论参考与方法依据。 In view of the problems such as high labor intensity of workers,poor consistency of assembly results and low assembly efficiency in the current bolt assembly in scenarios with uncertain postures,aiming at the industrial scenario of disassembling and assembling bolt connections during the roll changing process of steel rolling mills in steel mills,integrating industrial robots,stereoscopic vision positioning and force control technologies,a robot bolt assembly method based on point cloud and six-dimensional force information is proposed.Through a visual positioning strategy guided by vision and an adaptive straight-through filtering algorithm based on mode,the filtering threshold is dynamically adjusted to eliminate the interference of environmental noise,and the pose information of the target hole and the bolt head is obtained.On this basis,bolt assembly based on force information is carried out and a"rice"shaped variable step size search strategy is proposed to realize that the bolt end face is included by the target hole.A admittance control strategy was adopted to achieve the compliant insertion of bolts into the target hole.A hole center positioning method based on the vertical diameter theorem is proposed,achieving the coincidence of the bolt axis and the target hole axis.This method breaks through the dependence of traditional teaching programming on structured environments and provides an extensible technical framework for intelligent assembly in complex industrial scenarios.The results show that the success rate of bolt assembly by the proposed method in unknown scenarios is greater than 90%,and the time consumption is less than 5 seconds.This method effectively realizes bolt assembly in unknown environments.Provide theoretical references and methodological basis for the automation of bolt assembly.
作者 王国营 丁发 周玉朋 李素芳 常青 张付祥 李俊慧 Wang Guoying;Ding Fa;Zhou Yupeng;Li Sufang;Chang Qing;Zhang Fuxiang;Li Junhui(Shijiazhuang Iron and Steel Co.,Ltd.,Shijiazhuang 050031,Hebei;Hebei Xuanhua Construction Machinery Development Co.,Ltd.,Zhangjiakou 075105,Hebei;HBIS Group Co.,Ltd.,Shijiazhuang 050023,Hebei;School of Mechanical Engineering,Hebei University of Science and Technology,Shijiazhuang 050018,Hebei)
出处 《河北冶金》 2025年第6期89-96,共8页 Hebei Metallurgy
基金 河北省重大科技支撑计划项目(242Q1802Z) 河北省十大领域“机器人+”应用“揭榜挂帅”项目 石家庄市驻冀高校产学研合作项目(241080507A)。
关键词 螺栓装配机器人 立体视觉 位姿估计 六维力传感器 力控策略 bolt assembly robot stereo vision pose estimation six-dimensio
  • 相关文献

参考文献4

二级参考文献21

共引文献7

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部