基于优化位姿集的工业机器人运动学参数辨识方法研究  

Research on kinematics parameter identification method of industrial robot based on optimized pose set

作　　者：张颖 乔贵方[1,2] 王保升[3] 刘娣[1] 田荣佳 ZHANG Ying;QIAO Guifang;WANG Baosheng;LIU Di;TIAN Rongjia(School of Automation,Nanjing Institute of Technology,Nanjing 211167,China;School of Instrument Science and Engineering,Southeast University,Nanjing 210096,China;Research Department of Intelligent Manufacturing Equipment,Nanjing Institute of Technology,Nanjing 211167,China)

机构地区：[1]南京工程学院自动化学院,江苏南京211167 [2]东南大学仪器科学与工程学院,江苏南京210096 [3]南京工程学院智能制造装备研究院,江苏南京211167

出　　处：《中国测试》2023年第3期91-95,103,共6页China Measurement & Test

基　　金：国家自然科学基金项目(5190525);中国博士后科学基金(2019M650095);南京工程学院校级科研基金项目(CKJB202104)。

摘　　要：针对在高端制造领域工业机器人绝对定位误差仍无法满足精度要求的问题,提出一种基于优化位姿集的机器人定位精度提升方法。首先,基于M-DH模型对待标定机器人Staubli TX60建立运动学模型,并基于位姿微分变换方法构建该机器人的运动学误差模型;其次,利用IOOPS算法优化筛选机器人的辨识位姿集;最后,提出一种基于PSO-LM优化算法的运动学参数辨识方法,并通过实验验证运动学参数辨识精度。实验结果表明:基于PSO-LM混合优化算法辨识后的TX60机器人的平均综合位置/姿态误差分别从(0.5777 mm,0.0039 rad)降低为(0.0816 mm,0.0014 rad)。该文提出的PSO-LM混合优化算法具有较好的辨识精度和收敛速度,并且基于优化辨识位姿集获取的运动学模型具有更好的泛化能力。A method to improve the positioning accuracy of robots based on the optimized pose set is proposed,aiming at the problem that the absolute positioning error of industrial robots in the high-end manufacturing field still cannot meet the accuracy requirement.Firstly,the kinematic model of the robot Staubli TX60 is built based on the M-DH model,and the kinematic error model is built based on the pose differential transformation.Secondly,the identification pose set of the robot is optimized and screened by using the IOOPS algorithm.Finally,a kinematic parameter identification method is proposed based on the PSO-LM optimization algorithm,and the accuracy of kinematic parameter identification is verified through experiments.The experimental results show that the average comprehensive position/attitude error of the TX60 robot after identification is reduced from(0.5777 mm,0.0039 rad)to(0.0816 mm,0.0014 rad)based on the PSO-LM hybrid optimization algorithm.The PSO-LM hybrid optimization algorithm proposed in this paper has better identification accuracy and convergence speed,and the obtained kinematic model based on the optimized identification pose set has better generalization ability.

关 键 词：工业机器人 运动学参数 粒子群算法 机器人标定 观测指数 

分 类 号：TP242.2[自动化与计算机技术—检测技术与自动化装置] TB9[自动化与计算机技术—控制科学与工程]