Kinematic Calibration in Local Assembly Space of a Six-axis Industrial Robot for Precise Assembly

The research on the calibration of industrial robots mainly focuses on the global workspace, but it is difficult to ensure that industrial robots have good absolute positioning accuracy in the workspace. This paper proposes a kinematic calibrating method of industrial robot in local assembly space to improve the positioning accuracy. The kinematic error model of industrial robot is established based on modified Denavit-Hartenberg (MDH) model. The influence of redundant error parameters on kinematic parameter identification is analyzed. The method used in kinematic parameters identification is improved by using correlation tolerance and matrix singular value decomposition. Then, simulation and experimental test are carried to investigate the performance of the calibrating method. The experimental results indicate that the positioning accuracy inside the workspace is significantly reduced from 1.716 mm to 0.149 mm.