Small-sample optimal quality control method based on process capability

Abstract

The invention discloses a small-sample optimal quality control method based on a process capability. The method includes the following steps: 1. determining an objective function; 2. determining constraint conditions; 3. determining a concrete form of a model. Based on Bayes correction, the method studies a small sample which has utilization values to the correction and screens and discriminates the small sample through construction of a rapid judgment model and judges a process quality problem and thus under a condition that the process quality is known, a process quality control chart is designed to monitor a production process and the control chart is continued to be corrected according to a produced small sample and finally a better control effect can be achieved finally. At the same time, setting of a benchmark table also provides a reference for judgment of the process capability problem through use of the small sample.

Description

technical field [0001] The invention relates to a small sample optimal quality control method based on process capability. Background technique [0002] At present, the use of small samples for process quality inspection has always been a key issue that many scholars are committed to solving. The reason why it is so valued is that it has very important practical application value in production. However, small samples are unstable, so it is difficult to give a specific and general theoretical method. The idea of ​​process quality management has a long history, and its core idea is to use process capability instead of unqualified rate to judge whether the product meets the production requirements. In the 1920s, American scholar Shewhart [1] proposed the concept of process control and the method of realizing process control, and drew a process quality control chart, which is the Shewhart control chart that is still widely used today. When Zhang Gongxu and Lu Chunshan [2-3] i...

AI Technical Summary

Problems solved by technology

Because the small sample itself has great volatility, this means that the inappropriate selection of small samples directly leads to large deviations in the results during correction, increasing the possibility of false alarms or false alarms

In the previous models, no explanation was given on how to take a small sample, and it was believed that the larger the sample, the better even though it was a small sample, while ignoring the value of small samples for correction

Images