Abstract: To solve the problems that the error model contained the product of the orientation errors and the position vectors, which affected the accuracy of parameter calibration and recognition, when the pose errors of robot end-effectors were represented in the basic frame. A pose error calibration model of robots based on the POE including joint constraints was established where the tool frame was used as the reference frame, so as to avoid the influences of the product of orientation errors and the position vectors on the accuracy of parameter calibration and recognition. The UR5 robot was used as the calibration object, and the LeciaAT960-MR laser tracker was used as the measuring equipment to perform the parameter calibration tests. The test results show that the mean value of norms of the position errors and orientation errors of UR5 robot after calibrations are reduced by 91.07% and 89.16% respectively.

Key words: robot, orientation error, parameter calibration, tool coordinate, product of exponential(POE)

摘要： 机器人末端执行器位姿误差在基础坐标系中表示时，误差模型中包含姿态误差与位置矢量的乘积项，影响了参数标定识别精度。以工具坐标系为参考系，给出一种基于指数积公式包含关节约束条件的机器人位姿误差标定模型，避免了姿态误差与位置矢量的乘积项对参数标定识别精度的影响。以UR5机器人为标定对象，采用LeciaAT960-MR激光跟踪仪为测量设备，进行参数标定试验。试验结果表明，经参数标定后UR5机器人位置误差模和姿态误差模的平均值分别减小了91.07%和89.16%。

关键词: 机器人, 姿态误差, 参数标定, 工具坐标系, 指数积