基于ZRM-MDH模型转换的串联机器人运动学参数标定

Kinematic Parameters Calibration Method of Serial Robot Based on ZRM-MDH Model Transformation

摘要:

运动学参数误差是影响工业机器人绝对定位精度的主要因素，通过误差标定能够有效地提高工业机器人的精度。运动学模型的完整性、连续性与冗余性对运动学参数的辨识精度影响较大。为尽可能地提高机器人的标定精度，并易于实现机器人误差补偿，本文提出一种基于ZRM-MDH模型转换的机器人运动学参数标定方法。首先，基于零参考模型(ZRM)建立TX60型串联工业机器人的位姿误差模型，结合测量位姿误差辨识ZRM的参数；其次，基于圆点分析法将ZRM转换成MDH模型。在TX60型机器人前侧工作区域内任意选择50个测量点，实施运动学参数误差标定。实验表明，基于MDH模型标定后的机器人平均综合定位误差为0.081mm，而经过ZRM-MDH模型转换后的机器人平均综合定位误差为0.062mm。为验证标定方法的稳定性，在TX60型机器人前侧工作区域内，选择5个区域实施运动学参数误差标定，结果表明，基于ZRM-MDH模型转换获得的标定精度稳定性相对较好。

Abstract:

Kinematic parameter error is the main factor which affects the absolute positioning accuracy of industrial robots. The accuracy of industrial robots can be effectively improved through error calibration. The completeness, continuity and redundancy of the kinematic model have great impacts on the identification accuracy of the kinematic parameters. To improve the accuracy of robot calibration and perform robot error compensation easily, a method of robot kinematic parameter calibration based on ZRM-MDH model transformation was presented. Firstly, the kinematic error model of the serial industrial robot TX60 was established based on the zero reference model (ZRM). The parameters of the ZRM modified OH mudel were identified with the measured pose error. Secondly, the ZRM was transformed into a MDH model through the method of circle point analysis. Totally fifty points were selected in the front workspace of robot TX60 for the kinematic parameter error calibration. The experimental results showed that the average comprehensive positioning error calibrated based on the MDH model was 0.081mm. The average comprehensive positioning error calibrated based on the ZRM-MDH model transformation was 0.062mm. To verify the stability of the calibration method, five areas were selected in the front workspace of robot TX60 for kinematic parameter error calibration. The experimental results showed that the calibration accuracy stability obtained based on ZRM-MDH model transformation was better. Therefore, the kinematics parameter calibration method proposed can effectively improve the accuracy of the robot calibration.