Industrial robots are getting widely applied due to their low use-cost and high flexibility. However, the low absolute positioning accuracy limits their expansion in the area of high-precision manufacturing. Aiming to improve the positioning accuracy, a compensation method for the positioning error is put forward in terms of the optimization of the experimental measurement space and accurate modelling of the positioning error. Firstly, the influence of robot kinematic performance on the measurement accuracy is analysed, and a quantitative index describing the performance is adopted. On this basis and combined with the joints motion characteristics, the optimized measurement space in joint space as well as Cartesian space is obtained respectively, which can provide accurate measurement data to the error model. Then the overall model of the positioning error is constructed based on modified Denavit–Hartenberg method, in which the geometric errors and compliance errors are considered comprehensively, and an error decoupling method between them is carried out based on the error-feature analyses. Experiments on the KUKA KR210 robot are carried out finally. The mean absolute positioning accuracy of the robot increases from 1.179 mm to 0.093 mm, which verifies the effectiveness of the compensation methodology in this article.

中文翻译：

---

考虑优化测量空间的工业机器人绝对定位误差补偿

工业机器人因其使用成本低、灵活性高而得到广泛应用。然而，低绝对定位精度限制了它们在高精度制造领域的扩展。以提高定位精度为目标，从优化实验测量空间和精确建模定位误差等方面提出了一种定位误差补偿方法。首先分析了机器人运动学性能对测量精度的影响，并采用了描述性能的量化指标。在此基础上，结合关节运动特性，分别得到关节空间和笛卡尔空间的优化测量空间，为误差模型提供准确的测量数据。然后基于改进的Denavit-Hartenberg方法构建定位误差的整体模型，综合考虑几何误差和柔量误差，并在误差特征分析的基础上对其进行误差解耦。最后在 KUKA KR210 机器人上进行实验。机器人的平均绝对定位精度从1.179 mm提高到0.093 mm，验证了本文补偿方法的有效性。