| 皇甫亚波，杭鲁滨，程武山，余 亮，李文涛，杨国彬，沈铖玮.基于LM算法的机器人运动学标定[J].轻工机械,2017,35(4):1-7 |
| 基于LM算法的机器人运动学标定 |
| Design of Far Infrared Low Temperature Thawing Device |
| |
| DOI：10.3969/j.issn.1005 2895.2017.04.001 |
| 中文关键词: 串联机器人 运动学标定 LM算法 DH参数运动学模型 激光跟踪仪 |
| 英文关键词:serial robot kinematic calibration LM(Levenberg Marquardt) algorithm DH（Denavit Harteriberg） parametric kinematics model laser tracker |
| 基金项目:国家自然科学基金(NFS51475050)；上海市科学技术委员会重点支撑项目(12510501100) |
|
| 摘要点击次数: 1865 |
| 全文下载次数: 2289 |
| 中文摘要: |
| 机器人参数估计与误差补偿的标定方法是提高机器人末端位置定位精度的重要手段。课题组建立UR3机器人DH参数运动学模型，基于机器人末端工作空间的出现频率概率分布确定机器人测量空间域；以激光跟踪仪测量空间域内的机器人末端位置，构造出机器人末端位置误差与几何参数集间最小二乘方程目标函数，利用自适应步长的LM(Levenberg Marquardt)算法对其进行迭代求解，进行参数估计得到几何参数标定值，以此修正机器人系统理论运动学模型结构参数集。最后对标定结果进行验证和分析，结果显示标定后最大误差降低了47.48%，平均误差降低了37.98%，均方根误差降低了40.40%，证明了该标定方法的有效性和可靠性。 |
| 英文摘要: |
| The robot calibration method of parameter estimation and error compensation is an important means to improve the positioning accuracy of robot end effectors. Denavit Harteriberg kinematic model of the UR3 robots was established by the research group. Measurement space was determined by frequency probability distribution of workspace of robot end effectors. The positions of the robot end effectors in measurement space were measured by laser tracker. The least square objective function between the position error of the robot end effectors and the geometric parameters was established, the iterative solution of which were made by using adaptive step size LM(Levenberg Marquardt) algorithm. The kinematic model of robot system was corrected by these parameters. The calibration results was analyzed and verified in the end. The result showed that the maximum/mean/ root mean square position errors were reduced by 47.48%/ 3798%/40.40% after calibration. The validity and reliability of the mentioned method was proved by experiment. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
