Manufacturing process multivariate quality diagnosis classifier based on statistical method

Abstract

Provided is a manufacturing process multivariate quality diagnosis classifier based on a statistical method. The original data of the quality characteristics in a manufacturing process is collected and preprocessed. A process analysis is made of the multivariate quality characteristics of key procedures. The stability and whether there is an abnormal phenomenon are determined according to the datarecorded in a control chart. The process anomaly source is found through a statistical method. In order to make the classicization result more accurate, class mean, chi-square value, the weight ratiobetween class mean and chi-square value, and a mean distance stability criterion are introduced. The process capability coefficient condition is strict. The determined state is accurate. The algorithm complexity is low. The processing is fast. Multivariate quality, a misjudgment factor and a principal component factor are integrated. The classifier is of higher applicability. Parameter processingis standard. Data processing is perfect. The misjudgment probability is reduced. The problems of data bias and unit non-uniformity are solved. The manufacturing process multivariate quality diagnosisclassifier is more accurate than a support vector machine. An anomaly diagnosis technology can be realized.

Description

technical field [0001] The invention relates to the technical field of quality diagnosis in the processing and manufacturing process of mechanical products, in particular to a multivariate quality diagnosis classifier in the manufacturing process based on statistical methods. Background technique [0002] The modern manufacturing process is multivariate and highly correlated, and the process monitoring of this kind of production process is called multivariate quality control (MQC) or multivariate statistical process control (MSPC). The process of finding the cause of the loss of control is known as MSPC diagnosis or anomaly identification. There are two main types of methods: one is statistical decomposition techniques; the other is techniques based on machine learning. Mainstream decomposition techniques include principal component analysis (PCA), feature space comparison method, MTY method, step-down method, and multi-directional kernel principal component analysis method...

AI Technical Summary

Problems solved by technology

However, these methods usually involve complex statistical procedures, which are not conducive to the application

Images