Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Bayesian framework-based dynamic soft measurement modeling method and device

A Bayesian framework and modeling method technology, applied in the direction of instruments, adaptive control, control/regulation systems, etc., can solve problems such as feedback results that affect measurement, deterioration of control performance, and reduction of model tracking capabilities

Active Publication Date: 2013-09-04
TSINGHUA UNIV
View PDF5 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] First: usually only process data points that match the timing of quality data points are considered. However, most of the actual industrial processing processes are in dynamic changes. A certain state of the system at the current moment must be determined by the state of the previous period of time. It is not determined by the state at a certain time point alone, so the measurement result is not accurate enough;
[0005] Second: ignoring the process delay or even considering the process delay factor, the method of simply using a certain process data point for modeling, due to the lack of description of the process dynamic characteristics, makes the applicability of the soft sensor model limited to Under steady-state conditions, once the operating point changes or a disturbance occurs, the tracking ability of the model will be greatly reduced
The used process data only accounts for a small part of the total data, and most of the data containing rich process information is directly discarded, so the measurement results are not accurate enough;
[0006] Third, since the quality variable of the static soft sensor model will change drastically when the secondary variable changes drastically, from the control point of view, when the output of the soft sensor is used as the feedback signal of the control loop, the control performance will have a greater magnitude deterioration, which ultimately affects the feedback results of the measurement

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
  • Bayesian framework-based dynamic soft measurement modeling method and device
  • Bayesian framework-based dynamic soft measurement modeling method and device
  • Bayesian framework-based dynamic soft measurement modeling method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0054] In this embodiment, the dynamic soft sensor modeling method based on the Bayesian framework includes the following steps:

[0055] Step A: Establish a first-order impulse response model and a support vector machine model;

[0056] The first-order impulse response model is:

[0057] x ( t ) = x 1 ( t ) x 2 ( t ) · · · x m ...

Embodiment 2

[0069] This embodiment embodies the step A on the basis of the first embodiment;

[0070] Step A1: Construct the formula about u as described in formula (1) and formula (3) k (...,t)(k=1,...,m) and the system structure relationship of y(t);

[0071] Step A2: Get and u k (...,t)(k=1,...,m) corresponding data U k (...,t)(k=1,...,m) and with U k (..., t) (k=1,...,m) is a data sample composed of Y(t) mapped to each other, and forms an increasing time series according to the sampling time; usually when collecting the U k (…,t)(k=1,…,m) follow Shannon’s theorem;

[0072] Step A3: Standardize the data in the data sample so that the data has zero mean and unit variance in each dimension;

[0073] Step A4: Set the initial parameter α of the first-order impulse response model and support vector machine model training k ,τ k , impulse response sequence length L, training parameters ε, C, RBF kernel parameter σ and iteration stop threshold μ:

[0074] Step A5: Put the U in the dat...

Embodiment 3

[0077] This embodiment further details step A on the basis of embodiment two.

[0078] Such as figure 2 As shown, Step S1: On the basis of the analysis of the production process mechanism or actual experience, construct a systematic structural relationship between a difficult-to-measure leading variable and multiple easy-to-measure auxiliary variables, namely

[0079] u k (...,t)(k=1,...,m) and y(t) are the system structure relationship; and the parameter constraint range of the system is given.

[0080] The system structure relationship is: the system output y(t) represents the unpredictable leading variable, and the system input u k (…,t)(k=1,…,m) represent relevant and easily measurable auxiliary variables. The system is composed of a dynamic link and a nonlinear static link in series, and the dynamic link is expressed in the form of a first-order impulse response model:

[0081] Among them, m is the dimension of the input variable and also the dimension of the state v...

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 a Bayesian framework-based dynamic soft measurement modeling method and a Bayesian framework-based dynamic soft measurement modeling device. The method comprises the following steps of: A, establishing a first-order impulse response model and a support vector model; B, measuring and inputting uk(ellipsis,t)(k=1,ellipsis,m), calculating a state variable x(t) and calculating y(t); and C, outputting the y(t).

Description

technical field [0001] The invention relates to the field of computers, in particular to a Bayesian framework-based dynamic soft sensor modeling method and device in industrial production processes. Background technique [0002] In the process of industrial production control, quality variables are usually composed of indicators related to the quality of the product, such as the dry point of diesel, the melt index of polypropylene, the pH value in the chemical reactor, etc. However, quality variables usually cannot be measured online, and can only be obtained through several hours of laboratory analysis, and the laboratory test data is too late to be used as a feedback signal for the control loop. High-quality control requires Not met; even though there are alternative online analytical instruments for some quality variables, these are often expensive and post-maintenance. Based on the above reasons, in the fields of chemical process, biopharmaceutical, steel forging, etc.,...

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): G05B13/04
Inventor 黄德先尚超高莘青吕文祥
Owner TSINGHUA UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com