Method of Postures Selection for Industrial Robot Joint Stiffness Identification

Abstract

Industrial robots are being widely applied to machining operations, and are gradually becoming competitive with traditional CNC machining centers. Obtaining accurate stiffness values of robotic joints is the foundation for deflections compensation in case of large cutting forces. A number of factors influence the accuracy of joint stiffness identification, especially robotic posture. This paper proposes a robust and accurate method for selecting suitable postures in the joint stiffness identification. The identification process of the joint stiffness matrix is presented, an index considering both the dexterity and the condition number of the observation matrix is then developed, and the procedure for postures selection based on it is provided. The results of simulations and experiments show that the proposed method is more robust and accurate than classical method.