Uncertainty Models And Its Evaluation Methods Of Dimension Measurement Results For CMM

Evaluating the uncertainties of the measurement results properly for Coordinate Measuring Machine (CMM) is difficult because the measurement tasks of CMM are varied and the measurement errors are complex, which has limited the application of CMM. Task-oriented uncertainty evaluation has become the urgent problem needed to be solved in the application of CMM.In the thesis, the CMM dimension measurement is taken as an object. The uncertainty analyses, uncertainty models, uncertainty estimation methods of CMM task-oriented measurement are studied systematically. The motivation and innovation of the research works are as follows.(1) The measurement uncertainty analyses are studied systematically. The traditional error-source analysis method is not suitable for the uncertainty analysis of CMM task-oriented measurement. An uncertainty analysis method based on the value properties of the measurement system is presented. The uncertainty analysis is integrated with the statistical analysis for the measurement system in the proposed method. The method is suitable for all kinds of measurement models, measuring methods and measuring instruments. This uncertainty analysis method is simple, valid and practical, which is important for the application and popularization of measurement uncertainty.(2) The uncertainty models of CMM task-oriented measurement are established. The concrete modeling methods about the uncertainty models of transparent box and black box are provided. Based on virtual coordinate measuring machine, a simple and practical uncertainty evaluation method, SVCMM, is proposed. The uncertainty model of transparent box is based on measurement model, and the uncertainty model of black box is based on value analysis. The applicability and limitation of transparent-box uncertainty model are analyzed, and the study is focused on the application of black-box uncertainty model for CMM task-oriented uncertainty evaluation.(3) The uncertainty estimation methods are studied systematically. The sensitivity analysis method, Monte Carlo Method (MCM) and SVCMM are compared. A method based on the optimal combination of uncertainty value analysis, black-box uncertainty model, MCM simulation and statistical parameter estimation is proposed to evaluate task-oriented uncertainties of CMM measurement results, which is important for evaluating the uncertainties of measurement results scientifically and reasonably.(4) Based on the above uncertainty theories and methods, the uncertainty evaluation of three typical dimension measurement schemes of CMM is studied. The relationship between the measurement scheme and the measuring accuracy is analyzed according to the uncertainty evaluation results of the three measurement schemes. It provides a reliable theoretical basis for improving the application value of CMM.(5) An offline uncertainty evaluation software for CMM dimension measurement is developed. The software can realize automatic evaluation, and give out the evaluation reports according with Geometrical Product Specifications (GPS). It provides a platform for evaluating the uncertainties intelligently of CMM measurement results.(6) Based on the evaluation examples, the consistency and diversity of the different evaluation methods are discussed, and the characteristics and applicability of the methods are summarized. The evaluation scheme of using value analysis method to establish black-box uncertainty model and combining MCM application in the process of evaluation is proved to be optimal. The research work can provide a reference for improving the application value of CMM.