Cross-correlation-based feature construction method for production process data modeling

A production process, data modeling technology, applied in data processing applications, character and pattern recognition, instruments, etc., can solve problems such as not very comprehensive, and achieve the effect of improving the prediction accuracy

Pending Publication Date: 2020-02-28
INSPUR ARTIFICIAL INTELLIGENCE RES INST CO LTD SHANDONG CHINA
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The method used is generally the addition, subtraction, multiplication and division of the original data, or polynomial features, which is not very comprehensive

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Cross-correlation-based feature construction method for production process data modeling
  • Cross-correlation-based feature construction method for production process data modeling

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment

[0035] For example, a cross-correlation operator with five rows and three columns (usually the shape of the cross-correlation operator is an odd number for easy calculation) is as follows:

[0036]

[0037] Observing the digital characteristics of this cross-correlation operator, it can be found that this is a Gaussian-like distribution with a high middle and low surroundings. Since the center of the cross-correlation operation falls on the current data point of interest, the significance of this design is in the time series The weights are assigned according to the sequence, and the weights are also assigned in the feature dimension, and such a 5x3 area is integrated to extract features.

[0038] If the currently collected samples focus more on the correlation between variables, then the cross-correlation operator can be set with a larger number of columns, such as the transposition of the operator mentioned above.

[0039] If you want to consider the correlation between t...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides a cross-correlation-based feature construction method for production process data modeling. The cross-correlation-based feature construction method is used for product quality prediction in the field of intelligent manufacturing. The method comprises the following steps: normalizing and preprocessing collected production process data; setting hyper-parameters of a cross-correlation operator according to the data characteristics and the task purpose; setting the weight of a cross-correlation operator; performing operation on the data matrix by using a set cross-correlation operator, and performing sliding calculation on the cross-correlation operator in an interested data area to obtain a cross-correlation operator; obtaining a feature matrix after all regions of interest are operated by using a defined cross-correlation operator; connecting the feature matrixes in series to obtain a new feature matrix set. The obtained feature matrix can be used for anomaly detection, classification, regression and other tasks of machine learning, and then a predictive maintenance result is obtained.

Description

technical field [0001] The invention relates to a feature construction method, in particular to a cross-correlation-based feature construction method for production process data modeling, and belongs to the technical fields of production process modeling, cross-correlation, machine learning, and feature construction. Background technique [0002] The final quality of the product can often be reflected by the data collected during the production process. Isomorphic data analysis can play a role in predicting the quality of final products. However, how to extract good features for prediction is a very important technical issue. The data analysis work of the production process is mainly realized through data collection, machine learning modeling and computing power support. The data is generally stored with the MES system or other data storage devices, and after processing and processing, it forms useful data that can be used, and these data express various elements in the pr...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06Q10/06G06K9/52
CPCG06Q10/06395G06V10/42Y02P90/80
Inventor 段强安程治李锐金长新
Owner INSPUR ARTIFICIAL INTELLIGENCE RES INST CO LTD SHANDONG CHINA
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap