Accuracy Improvement for Industrial Robot Based on Joint Position Sensitive Virtual Tool Transformation Fitting Method

Abstract

As industrial robots wildly applied in advanced manufacturing fields, the accuracy requirement of robot keeps increasing. In this paper, a joint position sensitive virtual tool transformation fitting method is proposed to improve the pose accuracy of industrial robot. Firstly, the kinematic model based on Modified Denavit-Hartenberg (MD-H) model and the joint position sensitive error model are presented. Secondly, a virtual tool transformation fitting method based on Chebyshev interpolation polynomial is proposed. Finally, several experiments have been done to validate the performance of the proposed method. The experimental results show that the average position and orientation errors are reduced from (0.61883mm, 0.00418rad) to (0.06660mm, 0.00078rad). The maximum position and orientation errors decrease to (0.23912mm, 0.00263rad). The proposed method demonstrates significant advantages in terms of improvement balance between position and orientation accuracy.