Research On Error Analysis And Correction Technology Of Articulated Arm Coordinate Measuring Machine

With the continuous advancement of modern technology,the machinery manufacturing industry is moving toward large-scale integration,automation,in addition,higher and higher requirements are placed on measurement technologies and equipment.As far as the field of precision measurement is concerned,the bloated structure and expensive manufacturing cost of conventional large-scale coordinate measuring machines make it far from the modern requirements of the current industry.Therefore,in order to achieve rapid and accurate measurement of high mobility and high flexibility in the process of product manufacturing,the articulated arm coordinate measuring machine came into being.Thus,the research of this project has important theoretical significance and engineering value.This paper takes the Hexagon ROMER 6-DOF articulated arm coordinate measuring machine as the research object,analyzes its mechanical structure and performance characteristics,discusses the basis and principle of the freedom degree selection in detail,and using the D-H matrix method to establish a matrix mathematical model based on the forward kinematics rule.Then,based on the numerical substitution and Monte Carlo method,the measurement space of the probe movement is solved and simulated.The error source categories of the articulated arm coordinate measuring machine are studied,and four kinds of systematic errors are explained in detail.Under the premise of exploring the calibration principle of the structural parameters,a structural parameter calibration strategy based on the elimination of redundant least squares method is designed.Analyzed the RBF neural network prediction algorithm and formulate a correction idea for joint rotation angle error.Discussed the difficulty in correcting the measuring arm's force deformation error and the whole machine's thermal deformation error,then the general steps of calibration are given in turn.Finally,using MATLAB to simulate the two critical system errors.It is confirmed that the elimination of redundant least squares method is truly effective for the structural parameters calibration of the articulated arm coordinate measuring machine,and at the same time,the measurement error of the extremity probe iscorrected.In addition,the joint angle predicted by the trained RBF neural network was compared with the analog angle encoder readings,and the results showed that the accuracy of the predicted angle was improved.