Active enhanced soft measurement method based on sample expansion and screening

A sample expansion and soft-sensing technology, applied in the field of soft-sensing, can solve problems such as shortage and poor reliability of soft-sensing models, and achieve the effect of improving prediction accuracy

Pending Publication Date: 2021-11-12
ZHEJIANG UNIV OF TECH
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problem of poor reliability of the soft sensor model caused by the lack of original labeled samples in

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
  • Active enhanced soft measurement method based on sample expansion and screening
  • Active enhanced soft measurement method based on sample expansion and screening
  • Active enhanced soft measurement method based on sample expansion and screening

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0068] The present invention will be further described below in conjunction with the accompanying drawings.

[0069] refer to Figure 1 ~ Figure 4 , an active enhanced soft-sensing method based on sample expansion and screening, the specific steps are as follows:

[0070] (1) Obtain industrial polyethylene melt index data

[0071] The data of this process comes from the industrial polyethylene process of a factory. In this process, the melt index (Melt Index, MI) index is usually used to measure the quality of polyethylene products. However, the melt index cannot be measured with an on-line sensor, and can only be obtained through off-line experimental analysis, which will cause a large time lag and produce substandard products. At the same time, the polyethylene production process frequently changes its production conditions and product formula ratios to produce different types of products, resulting in the lack of collected labeled data at the beginning of the operation a...

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 discloses an active enhanced soft measurement method based on sample expansion and screening. The method comprises the following steps: 1) obtaining industrial polyethylene melt index data; 2) dividing a data set of polyethylene data and preprocessing the polyethylene data; 3) generating virtual samples and expanding the data set; 4) establishing a selective generative adversarial network based on support vector regression (SGAN-SVR) soft measurement model; and 5) performing model performance evaluation. According to the method, a high-quality generated sample is screened from the virtual samples generated by a Wasserstein GAN with gradient penalty (WGAN-GP) model according to a centroid measurement criterion and statistical characteristics of original samples, and is used as a supplement of the original samples, so that the prediction accuracy of a support vector regression (SVR) model on an MI is improved.

Description

technical field [0001] The invention relates to the technical field of soft sensing, in particular to an active enhanced soft sensing method based on sample expansion and screening. Background technique [0002] In recent years, with the wide application of the Internet, Internet of Things and data acquisition and storage systems in the industrial field, massive amounts of data have been collected and recorded. However, we must notice that there is a problem of lack of original labeled samples in the context of big data. Although the amount of data obtained in many process industries is huge, many of the data are repeated samples, and the useful information is very limited. In addition, data are also limited due to the high cost of data acquisition or low incidence. Such a small number of useful samples cannot completely cover the entire effective space, the information is incomplete, and the ability to describe the original sample feature space is also insufficient. [0...

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
IPC IPC(8): G06K9/62G06N3/04G06N3/08G06N20/10
CPCG06N20/10G06N3/04G06N3/08G06F18/214
Inventor 刘毅戴云余清
Owner ZHEJIANG UNIV OF TECH
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