A novel calibration method for multi-robots system utilizing calibration model without nominal kinematic parameters

• Published in: Precision engineering 2017-10, Vol.50, p.211-221
• Main Authors: Qiao, Yu, Chen, Youping, Chen, Bing, Xie, Jingming
• Format: Article
• Language: English
• Subjects: Adjoint transformation; Calibration model; Kinematic parameters; Multi-robots system; Product of exponentials formula
• ISSN: 0141-6359; 1873-2372
• DOI: 10.1016/j.precisioneng.2017.05.009

•A novel method for multi-robots system calibration without nominal kinematic parameters is proposed.•The proposed model merely depends on the reference configuration rather than base frame and reference configuration.•The parameters in our model are totally independent, and the total number of the parameters is 4r+2p.•The proposed method avoids the operations of normalization and orthogonalization which are necessary in the POE model.•The kinematic parameters of a robot and the exact pose of the robot base frame relative to the unified frame can be obtained simultaneously. In this manuscript, a novel method for multi-robots system calibration utilizing a calibration model based on the Product of Exponentials formula (POE formula) without nominal kinematic parameters is proposed. The calibration model is characterized by the adjoint transformation which is adopted to represent the transformation relationship between two twists. Compared with the traditional POE formula model for calibration, the proposed model merely depends on the reference configuration rather than base frame and reference configuration. Then a coordinate which describes the robot is fixed and the end-effector pose with respect to the predetermined coordinate is measured several times. Least square method is adopted to identify the exact POE formula model with respect to the predetermined coordinate. The results of simulation and experiment reveal that 1) with the proposed model, nominal kinematic parameters are not required to establish the initial kinematic equation, 2) the result of the calibration is not only identifying the exact kinematic parameters of the robot but also determining the pose of the robot base frame with respect to a predetermined coordinate, 3) the calibration method can increase the accuracy of multi-robots system.