Research Of Kinematic Calibration For Robot Palletizer

With the continuous development of modern industry,palletizing technology as an important part of robot technology is widely used in all walks of life,which puts forward higher requirements on the positioning accuracy of robot palletizer.In this paper,the kinematic calibration method of self-developed robot palletizer was studied in order to improve its absolute accuracy.The main research contents are as follows:Firstly,the mechanism of error generation and error distribution of robot palletizer were deeply analyzed based on the two major evaluation indicators of robot positioning accuracy: repeatability and absolute accuracy,and on this basis,the error factors were classified.The graded mixing error compensation method combining the characteristics of error model and non-model calibration method was proposed aiming at the main error factors: geometric error and waist joint transmission error.The calibration methods based on error model is used for geometric error of robot palletizer.The parallel four-connecting rods error model of robot palletizer was built by loop incremental method,and based on this,a kinematic error model of the whole robot is established based on DH method.The number of geometric error was reduced to 8 by analyzing the sensitivities of geometric error parameters on positioning accuracy of robot,which can promote the efficiency of identification.The parameter identification model based on distance error and joint space error compensation method were used,and the effectiveness of the calibration model was verified by a numerical example.The non-model calibration method based on linear interpolation is adopted for the waist joint transmission error.The calibration of this error was divided into node division,error measurement,error estimation based on linear interpolation and error compensation in four stages by drawing on the basic idea of error compensation method based on spatial interpolation.With laser tracker as a measuring instrument,the calibration experiments were carried out by using graded mixing calibration method,circle points analysis,calibration method based on error model and non-model in view of spatial interpolation,and the compensation effects of different calibration methods were compared and analyzed.The experimental results shows that graded mixing calibration method improves the absolute accuracy of robot further 4.76% compared with error model calibration method,and the calibration efficiency is higher than that of the non-model calibration method,which has good effect in error separation,reliability and generality.