Self-calibration Method for Robot Based on Geometric Constraints

Focusing on the self-localization problems of the seeds sampling robot system during the operation process, this paper proposes a self-calibration method of industrial robot. It is based on geometric constraints of points in the coordinate system of the seed sampling box, which realizes the automatic acquisition of the absolute coordinates of robot target points by using the homogeneous coordinate transformation matrix and solves the self-localization problems. The test results show that the absolute value of the relative difference of theoretical coordinates and the actual coordinates is less than 5%, which means the method is correct and practicable and can meet the requirements of practical work. It is easy to implement and it has advantages of low cost and high accuracy, which is of great significance to solve the self- localization of the workpiece.