摘要
针对机器人装配过程难以建立精确的物理模型,导致机器人装配灵活性差、装配成功率低的问题,提出一种基于过程建模的机器人分阶引导装配方法,建立基于装配各阶段的力、力矩、位姿装配过程模型,结合装配状态与机器人执行动作之间的映射,形成机器人装配阶段引导策略。通过在所搭建的KUKA机器人装配平台上进行算法验证表明,装配阶段的识别准确率达到99%以上,分阶引导策略下机器人的装配成功率达到97.2%。
The accurate physical model is difficult to build in robot assembly,which leads to poor flexibility and low success rate of assembly.To solve this problem,a robot phased guided assembly method based on process modeling was proposed.An assembly process model based on force,torque,position and posture in each stage of the assembly was established.By combining with the mapping relationship between each stage in the assembly process and the robot s execution actions,a phased guiding strategy of robot assembly was formed.Through the built KUKA robot assembly platform,the algorithm verification showed that the recognition accuracy rate of each assembly stage reached more than 99%,and the assembly success rate under the phased guidance strategy reached 97.2%.
作者
杨旭亭
王孜悦
李凤鸣
宋锐
YANG Xuting;WANG Ziyue;LI Fengming;SONG Rui(School of Control Science and Engineering,Shandong University,Jinan 250100,China;Engineering Research Center of Ministry of Education for Intelligent Unmanned Systems,Jinan 250100,China;College of Science,Guilin University of Technology,Guilin 541006,China)
出处
《计算机集成制造系统》
EI
CSCD
北大核心
2021年第8期2321-2330,共10页
Computer Integrated Manufacturing Systems
基金
广东省重点领域研发计划资助项目(2020B090925001)
国家自然科学基金面上资助项目(61973196)
山东省重大科技创新工程资助项目(2019JZZY010430)。
关键词
机器人
装配建模
接触状态
分阶引导
robot
assembly modeling
contact state
stage guidance
作者简介
杨旭亭(1994-),男,河北沧州人,硕士研究生,研究方向:工业机器人柔性装配作业,E-mail:18202545015@163.com;通讯作者:王孜悦(2000-),女,山东济南人,本科生,研究方向:信息与计算科学,E-mail:wziyins27@163.com;李凤鸣(1984-),女,山东淄博人,博士研究生,研究方向:机器人柔性作业、机器人复杂作业技能学习,E-mail:lfmleg@foxmail.com;宋锐(1975-),男,山东济南人,教授,博士,博士生导师,研究方向:控制理论与应用、机器人复杂作业技能学习与优化,E-mail:rsong@sdu.edu.cn。
