Calibration of Robot Reference Frames for Enhanced Robot Positioning Accuracy

Abstract

Industrial robot manipulators are important components of most automated manufacturing systems. Their design and applications rely on modeling, analyzing, and programming the robot tool-center-point (TCP) positions with the best accuracy. Industrial practice shows that creating accurate robot TCP positions for robot applications such as welding, material handling, and inspection is a critical task that can be very time-consuming depending on the complexity of robot operations. Many factors may affect the accuracy of created robot TCP positions. Among them, variations of robot geometric parameters such as robot link dimensions and joint orientations represent the major cause of overall robot positioning errors. This is because the robot kinematic model uses robot geometric parameters to determine robot TCP position and corresponding joint values in the robot system. In addition, positioning variations of the robots and their end-effectors also affect the accuracy of robot TCP positions in a robot work environment. Model-based robot calibration is an integrated solution that has been developed and applied to improve robot positioning accuracy through software rather than changing the mechanical structure or design of the robot itself. The calibration technology involves four steps: modeling the robot mechanism, measuring strategically planned robot TCP positions, identifying true robot frame parameters, and compensating existing robot TCP positions for the best accuracy. Today, commercial robot calibration systems play an increasingly important role in industrial robot applications because they are able to minimize the risk of having to manually recreate required robot TCP positions for robot programs after the robots, end-effectors, and fixtures are slightly changed in robot workcells. Due to the significant reduction of robot production downtime, this practice is extremely beneficial to robot applications that may involve a rather large number of robot TCP positions. This chapter provides readers with methods of calibrating the positions of robot reference frames for enhancing robot positioning accuracy in industrial robot applications. It is organized in the following sections: Section 2 introduces basic concepts and methods used in modeling static positions of an industrial robot. This includes robot reference frames, joint parameters, frame transformations, and robot kinematics. Section 3 discusses methods and techniques for identifying the true parameters of robot reference frames. Section 4 presents applications of robot calibration methods in transferring existing robot programs among