伪目标迭代生成的机器人误差补偿算法
Robot Error Compensation Algorithm by Pseudo Target Iterative Generation
叶伯生 1金雄程 1黎晗 1邵柏岩 1李晓昆 1李思澳1
作者信息
- 1. 华中科技大学机械科学与工程学院,武汉,430074
- 折叠
摘要
针对机器人误差模型建立后的误差补偿问题,提出一种改进的伪目标迭代算法.该算法用每次迭代中生成新的伪目标来修正关节角,从而不断减小机器人实际误差.完善了伪目标迭代算法流程,提出了 5种新的不同的伪目标生成方法,分析了各种方法的特点和适用场景.结合多种伪目标生成方法提出了一种集成算法,进一步提高误差补偿精度.使用HSR-JR612机器人进行仿真实验,仿真结果表明,算法耗时在毫秒级别,补偿效果好,集成算法能进一步提高位姿补偿效果.最后使用UR10机器人与激光跟踪仪进行实验,实验结果表明,补偿后机器人末端位置误差可以减小到0.06 mm以内,姿态误差可以减小到0.025°以内.
Abstract
Aiming at the problems of error compensation after robot error modeling,an improved algorithm was proposed based on pseudo target iteration.This algorithm generated a new pseudo tar-get in each iteration to correct the joint angles,thereby continuously reducing the actual errors of the robot.The pseudo target iterative algorithmic process was improved to mature the algorithm flow,5 new and different false target generation methods were proposed,and the characteristics and applica-ble scenarios of each method were analyzed.An ensemble algorithm was proposed combining multiple false target generation methods to further improve the accuracy of error compensation.Using the HSR-JR612 robot to carry out simulation experiments,the simulation results show that the algorithm takes milliseconds and the compensation effectiveness is good,and the integrated algorithm may fur-ther improve the pose compensation effect.Finally,the UR10 robot and the laser tracker were used for experiments.The experimental results show that after compensation,the position errors of the ro-bot end may be reduced to less than 0.06 mm,and the attitude errors may be reduced to less than 0.025°.
关键词
机器人/误差补偿/定位精度/伪目标Key words
robot/error compensation/positioning accuracy/pseudo target引用本文复制引用
基金项目
湖北省重点研发计划(2021BAA197)
出版年
2024
国家科技期刊平台
NSTL
维普
万方数据