高级检索
洪磊, 杨小兰, 王保升, 吕东升. 基于球面均匀分布的焊接机器人TCP标定方法[J]. 焊接学报, 2020, 41(8): 14-21. DOI: 10.12073/j.hjxb.20191212001
引用本文: 洪磊, 杨小兰, 王保升, 吕东升. 基于球面均匀分布的焊接机器人TCP标定方法[J]. 焊接学报, 2020, 41(8): 14-21. DOI: 10.12073/j.hjxb.20191212001
HONG Lei, YANG Xiaolan, WANG Baosheng, LV Dongsheng. TCP calibration method based on spherical uniform distribution for welding robot[J]. TRANSACTIONS OF THE CHINA WELDING INSTITUTION, 2020, 41(8): 14-21. DOI: 10.12073/j.hjxb.20191212001
Citation: HONG Lei, YANG Xiaolan, WANG Baosheng, LV Dongsheng. TCP calibration method based on spherical uniform distribution for welding robot[J]. TRANSACTIONS OF THE CHINA WELDING INSTITUTION, 2020, 41(8): 14-21. DOI: 10.12073/j.hjxb.20191212001

基于球面均匀分布的焊接机器人TCP标定方法

TCP calibration method based on spherical uniform distribution for welding robot

  • 摘要: 当前焊接机器人工具中心点(TCP, tool center point)采用固定参考点法标定时,存在机器人位姿选择的随机性和分布的不均匀性,为解决这一问题,提出了一种基于球面均匀分布的TCP标定方法. 以机器人自带的“六点法”为初步标定基础,创建初始测量点位形;在离线仿真环境下,采用力学斥力迭代法构建以固定参考点为球心呈球面均匀分布的虚拟点,逐组计算使虚拟机器人第六轴末端中心处于各虚拟点处,剔除其中关节角超限、连杆之间发生碰撞的情形;最后调节实际机器人到筛选后的各测量点位形,应用最小二乘球面拟合法求解最终的TCP标定结果. 结果表明,该方法使机器人姿态在各测量点绕固定参考点均匀分布,最大限度增大了各测量点之间机器人位姿的差异度,可有效提高标定精度和稳定性.

     

    Abstract: In the current fixed reference point method for calibration of welding robot tool center point (TCP), there are problems of randomness of robot posture selection and non-uniformity of distribution. For this reason, a method of TCP calibration based on spherical uniform distribution is proposed. Based on the “six-point method”, the initial measurement configuration is created at first. Then, in the off-line simulation environment, the mechanical repulsion iteration method is used to construct the virtual points with spherical uniform distribution with the fixed reference point as the center of the sphere. The end center of the virtual robot sixth axis was located at each virtual point by group calculation, and the situations of joint angle over limit and link collision are eliminated. Finally, the actual robot is adjusted to the selected measurement configurations for the final TCP calibration, and the result is solved by the least square spherical fitting method. The results show that this method can make the robot posture evenly distributed around the fixed reference point at each measuring point, and maximize the difference degree of robot posture, it is verified that this method can effectively improve the calibration accuracy.

     

/

返回文章
返回