Processing quality monitoring method based on dynamic PCA-SVM

Abstract

The invention discloses a processing quality monitoring method based on a dynamic PCA-SVM (Principal Component Analysis-Support Vector Machine). The method includes: aiming at the time sequence correlation between non-uniformly sampled processing process data and quality data, extracting a time sequence dynamic relationship between the processing process data and the quality data by utilizing a dynamic time window; adopting a PCA method to carry out preprocessing and feature extraction such as denoising, dimension reduction and correlation elimination on related processing process variables, and improving the quality monitoring efficiency; aiming at poor quality monitoring accuracy, false alarm and missing alarm caused by a non-linear and high-dimensional complex association relationship between the processing process and the quality, taking principal component characteristics of the processing process and the quality data after dynamic PCA extraction as samples, training an SVM classifier, carrying out SVM model testing by using test samples, and finally, obtaining a machining process quality monitoring model with relatively high generalization. According to the method, the noise in the original process data is reduced, and the accuracy and the calculation efficiency of the quality monitoring model are further improved.

Description

technical field [0001] The invention belongs to the technical field of mechanical processing, and in particular relates to a processing quality monitoring method. Background technique [0002] The processing quality of complex mechanical products is usually determined by a number of quality characteristic factors that have certain correlations. In this multi-variable manufacturing process environment that affects processing quality, there is a correlation between many and complex quality-related data. , making the traditional process quality monitoring or analysis methods that assume variables are independent of each other difficult to work. The literature "Monitoring and Optimization of Processing Quality of Craft Products, Science Technology and Engineering, 2017, Vol17(18), p277-281" discloses a method for monitoring and optimizing process product quality. Before training the product processing quality monitoring optimization model, the method uses the principal componen...

AI Technical Summary

Problems solved by technology

However, for complex and dynamic processing processes, the processing data and quality data are asynchronously sampled, and there is a time series correlation between different processing data and quality data, it is difficult to extract the relationship between the two directly using the principal component analysis based on the variance maximization criterion. The time series dynamic relationship between them, the quality analysis and monitoring of the complex dynamic processing process has poor accuracy and low efficiency

Images