Dynamic evaluation method for roundness uncertainty based on Bayesian principle

Abstract

The invention provides a dynamic evaluation method for roundness uncertainty based on the Bayesian principle. The method comprises the steps: carrying out the roundness sampling on a measured object,and obtaining a group of measurement point data; fitting a circle according to the measurement point data, and further calculating a roundness error; collecting groups of measurement point data, calculating roundness errors of the groups of measurement point data, and taking the roundness errors as a group of random variables; establishing a random variable probability density function, and constructing constraint conditions of probability density; estimating a probability density function of the group of random variables by taking a sample origin moment as a condition and taking the probability density constraint condition as an objective function, and taking the probability density function as prior distribution; and acquiring another group of random variables, calculating to generate another probability density function, and fusing the probability density function of the another group of random variables into prior distribution to form posteriori distribution. The method can realizefusion of geometric tolerance measurement information, and achieves the dynamic evaluation of uncertainty.

Description

technical field [0001] The invention relates to precision measurement and computer application, in particular to a dynamic evaluation method of roundness uncertainty based on Bayesian principle. Background technique [0002] With the rapid development of industrial process informatization and intelligence, new manufacturing models such as Industry 4.0 and intelligent manufacturing have been widely discussed and studied. Under the new industrial manufacturing system, product measurement technology requirements that meet the characteristics of cyber-physical systems are necessary. According to modern error theory, when measuring the size of a workpiece, not only the size measurement result must be obtained, but also the uncertainty of the result must be included. In the new generation of product geometric specification (GPS), the measurement uncertainty is called execution uncertainty. degree into the system. [0003] The core of the Bayesian model is to realize the fusion o...

AI Technical Summary

Problems solved by technology

The existing research mainly includes the following contents: Bayesian method is adopted for the evaluation of measurement uncertainty of small samples in calibration and inspection, and each calibration or inspection data is effectively used, but the description of prior and posterior distributions is mainly based on Traditional statistical methods or empirical assumptions; a Bayesian measurement uncertainty evaluation method based on the maximum entropy theory is proposed, which realizes the maximum entropy expression form of the posterior distribution PDF, but its prior distribution is mainly based on empirical assumptions; using Based on the principle of Bayesian information fusion and statistical inference, a dynamic evaluation model of uncertainty has been established and verified by digital simulation experiments. Reflect the characteristics and advantages of Bayesian information fusion

In addition, the above studies are mainly aimed at the uncertainty evaluation of a single variable, and there are no relevant reports in the field of multivariate geometric tolerance measurement uncertainty

[0004] In view of the above problems, the research on measurement information fusion and dynamic evaluation methods for roundness error uncertainty is very limited, especially the field of information fusion does not involve the category of geometric tolerance

Images