Investigating the Posture Dependency on Positioning Error of a Six-Axis Industrial Robot
Tianhao Cui, Soichi Ibaraki
 
 
 
Abstract
 
For robotic machining, the performance of model-based offline compensation of the robot positioning error is pivotal. To enhance its performance, many scholars have investigated a calibrated kinematic model e.g. Denavit-Hartenberg (D-H) model or its enhancement by including one degree-of-freedom (DOF) stiffness parameter within the rotary plane. However, higher DOF errors, i. e. deviations of the rotary axis out of the rotary plane, are typically ignored for simplicity. In this paper, we proposed a novel measurement and modelling scheme that could quantify higher-order DOF positioning errors. The core of this method is that, instead of using a mechanistic model with the stiffness matrices, the influence of robot poses on the position and orientation of the rotating trajectory of each axis is directly observed by measuring the end effector trajectories of the robot using a laser tracker. Experiment results demonstrate the proposed scheme could significantly improve the prediction accuracy due to the inclusion of posture-dependent errors.
 
Keywords: Posture-dependent, industrial robot, calibration
 
 
 
 
|