直线电机驱动六自由度并联机器人运动学标定研究  

Kinematic Calibration of the 6-DOF Parallel Robot Driven by Linear Motor

作　　者：张鑫璐 刘志华[2] 蔡晨光[2] 孔明 符磊 ZHANG Xinlu;LIU Zhihua;CAI Chenguang;KONG Ming;FU Lei(Collage of Metrology&Measurement Engineering,China Jiliang University,Hangzhou,Zhejiang 310018,China;National Institute of Metrology,Beijing 100029,China;The Electrical Engineering College,Guizhou University,Guiyang,Guizhou 550025,China)

机构地区：[1]中国计量大学计量测试工程学院,浙江杭州310018 [2]中国计量科学研究院,北京100029 [3]贵州大学电气工程学院,贵州贵阳550025

出　　处：《计量学报》2025年第2期213-221,共9页Acta Metrologica Sinica

基　　金：国家自然科学基金面上项目(52075512);中央级公益性科研院所基本科研业务费专项资金(AKYZD2302)。

摘　　要：为了解决并联机器人运动学标定中由于位置和姿态量纲不统一而引起的寻优算法过程中迭代收敛速度慢问题,提出了对误差模型中位移和姿态函数数量级进行统一的方法。首先,以直线电机驱动的6-PSS并联机器人为研究对象,基于闭环矢量方程,通过降低在误差雅可比矩阵中位移函数和姿态函数的损失函数值构建无量纲化的运动学参数误差模型;其次,采用最小二乘算法辨识动静平台球铰铰链点、杆长、导轨方向向量在内的运动学参数误差,根据辨识结果进一步对控制模型的参数进行修正;最后,进行直线电机驱动的6-PSS并联机器人运动学标定实验,将AICON双目立体视觉装置放置于没有光线突变的室内环境中测量并联机器人末端执行器的实际位置和姿态,并将理论位置和姿态进行对比得到误差,进而对样机进行精度补偿。实验结果表明,经过运动学标定后,样机位置精度提升96.736%,姿态精度提升83.750%,精度补偿有效,验证了误差模型的有效性。To tackle slow iterative convergence in the optimization algorithm caused by inconsistent position and attitude dimensions in kinematic calibration of parallel robots,a method for unifying the order of magnitude of displacement and attitude functions in error models is proposed.First,taking the 6-PSS parallel robot driven by a linear motor as the research object,based on the closed-loop vector equation,a dimensionless kinematic parameter error model is constructed by reducing the loss function values of the displacement function and attitude function in the error Jacobian matrix.Second,the least squares algorithm is used to identify the kinematic parameter errors including the spherical hinge point,rod length,and guide rail direction vector of the dynamic and static platform,and the parameters of the control model are further modified based on the identification results.At last,a linear motor-driven 6-PSS parallel robot kinematics calibration experiment is conducted,and the AICON binocular stereo vision device is placed in an indoor environment without sudden changes in light to measure the actual position and attitude of the parallel robot’s end effector.The theoretical position and attitude are compared to obtain the error,and then the accuracy of the prototype is compensated.The experimental results show that after kinematic calibration,the position accuracy of the prototype is improved by 96.736%,the attitude accuracy is improved by 83.750%,and the accuracy compensation is effective,which verifies the effectiveness of the error model.

关 键 词：并联机器人 6-PSS 误差模型 最小二乘法 运动学标定 

分 类 号：TB934[一般工业技术—计量学]