Roundness uncertainty evaluation method based on probability density function estimation

Abstract

The invention provides a roundness uncertainty evaluation method based on probability density function estimation, and the method comprises the steps: carrying out the roundness sampling of a measuredobject, and obtaining a group of measurement point data; fitting a circle according to the measurement point data, and calculating a roundness error; acquiring a plurality of groups of measurement point data to obtain a plurality of roundness errors, and taking the roundness errors as a group of random variables; establishing a random variable probability density function, a sample origin momentgenerated through random variable calculation replacing a theoretical origin moment for mathematical change, and constructing constraint conditions of probability density; taking the origin moment ofthe sample as a condition, taking a probability density constraint condition as an objective function, and obtaining a probability density function of the roundness error; and performing numerical integration on the probability density function to calculate the standard deviation of the probability density so as to realize uncertainty evaluation of roundness measurement. The roundness error measurement uncertainty evaluation method can realize roundness error measurement uncertainty evaluation of small samples, and has the characteristics of fast algorithm convergence and stable calculation numerical value.

Description

technical field [0001] The invention relates to the field of precision metrology and computer application, in particular to a method for evaluating roundness uncertainty based on probability density function estimation. Background technique [0002] As an important indicator for evaluating cylindrical parts, roundness error plays a very important role in precision measurement tools and high-precision error evaluation of cylindrical parts in the fields of machinery manufacturing, electric power, transportation, aerospace, and automatic testing. 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. In addition, the evaluation of the uncertainty of geometric errors must be carried out on the basis of ...

AI Technical Summary

Problems solved by technology

The core problem of this evaluation process is the multivariate optimization problem of PDF parameters under the constraints derived from the principle of maximum entropy. Generally, statistical learning methods such as gradient descent method, Newton method or quasi-Newton method are used, but these methods all require Considering the objective function derivative problem, the calculation is more cumbersome

In addition, the above studies are mainly aimed at the measurement uncertainty evaluation of a single variable, and have not been extended to the uncertainty evaluation of geometric errors.

[0005] In view of the above description, for small-sample measurement data of geometric tolerances such as roundness error, the research on uncertainty evaluation methods without distribution assumptions is very limited, especially in the process of introducing intelligent optimization algorithms.

Images