A soft measurement modeling method based on pattern perception dynamic variational auto-encoding regression

The MADVAR model solves the modeling challenges of multi-mode and dynamic characteristics in industrial processes, enables accurate online prediction of key quality variables, and improves the generalization ability of soft measurement models.

CN122388447APending Publication Date: 2026-07-14NANTONG VOCATIONAL COLLEGE
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610838394.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-06-11
Publication Date
2026-07-14

AI Technical Summary

Technical Problem

Existing soft measurement modeling methods for industrial processes are difficult to effectively capture multi-mode and dynamic characteristics. In particular, traditional methods struggle to balance the coupling between multi-mode operational characteristics and dynamic temporal characteristics in industrial processes, making it difficult to accurately obtain key quality variables.

Method used

A pattern-aware dynamic variational autoencoder regression (MADVAR) model is adopted. The latent features of dynamic data are mined through an encoder-decoder structure. A Gaussian mixture distribution is introduced, and the Wasserstein distance is used to constrain the reconstruction process of dynamic data to build a mapping relationship between latent features and key quality variables, thereby realizing online quality prediction.

Benefits of technology

It improves the ability to model multi-modal dynamic coupling features, enhances the prediction accuracy and stability of key quality variables, and enables feature extraction and quality prediction to be completed within the same probabilistic framework.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122388447A_ABST
    Figure CN122388447A_ABST
Patent Text Reader

Abstract

The application discloses a soft measurement modeling method based on mode perception dynamic variational autoencoding regression, and comprises the following steps: (1) acquiring three-phase flow process data with multiple mode dynamic characteristics; (2) data division and pretreatment operation; (3) establishing a mode perception dynamic variational autoencoding regression model, and realizing the prediction of an online quality variable; and (4) predicting a three-phase flow process pressure variable and performing model performance evaluation. The encoder-decoder network with a convolution-deconvolution structure learns the hidden layer feature representation of dynamic data, the multiple mode characteristics of the data are mined in the hidden space by using a Gaussian mixture distribution, the mapping relationship between the latent variable and the key quality variable is effectively learned, in addition, the reconstruction of dynamic multiple mode data is constrained by using a Wasserstein distance, and finally the purpose of improving the accuracy of key quality inference is achieved.
Need to check novelty before this filing date? Find Prior Art