Soft measurement method for on-line detection of components in special rectification process

A rectification process and soft-sensing technology, applied in the field of rectification process control, can solve problems such as measurement lag, too many controlled quantities, and difficult parameter adjustment of the model, so as to improve stability, remove the influence of algorithm stability, and improve automatic The effect of adaptability

Pending Publication Date: 2022-03-04
QINGDAO UNIV OF SCI & TECH
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] Due to the large number of controlled quantities and complex interactions between components in the special distillation process, component controllers must be used in many control strategies to meet product quality control requirements. However, online detection of product components in actual industrial processes is still There are many problems. Most of the existing measurement methods have many problems such as low accuracy, measurement lag, poor reliability, and high cost. The measurement results are difficult to use for component control. However, soft measurement technology uses other measurement information that is easy to obtain and realizes it through calculation. Estimating the measured variables, realizing the indirect measurement of component information,

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
  • Soft measurement method for on-line detection of components in special rectification process
  • Soft measurement method for on-line detection of components in special rectification process
  • Soft measurement method for on-line detection of components in special rectification process

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0030] according to figure 1 , 2 As shown, this embodiment provides a soft sensor method for on-line detection of components in a special rectification process, including the following steps:

[0031] Step 1. After collecting historical data and normalizing it, the principal variable is extracted through principal component analysis, and the principal variable is used as input to establish a random forest model to obtain the variable importance index, and the variable value with a large random forest importance index is selected as Input variables into the soft sensor model;

[0032] Among them, the obtained principal variable is decomposed into eigenvalues. The eigenvalue decomposition method is to randomly select the eigenvalues ​​to split the spanning tree, and count the sum of the Gini index decline degrees of the branch nodes in each tree to obtain the importance of the features, and randomly The forest uses a parallel bagging method. After the eigenvalues ​​are decompo...

Embodiment 2

[0042] In this example, a single-column rectification process for the industrial production of ethylbenzene and styrene is constructed through the chemical engineering simulation software Aspen Plus and Aspen dynamic. The process flow is as attached to the description. figure 2As shown, the rectification is 12500Kg / h, and the rectification components are: 58.43wt% ethylbenzene, 41.5wt% styrene and a small amount of tar. The rectifying tower overhead product is the ethylbenzene of 99wt%, the styrene of 99.7wt%, wherein

[0043] (1) The feed flow rate is adjusted by the flow controller;

[0044] (2) The top reflux tank and the liquid level at the bottom of the tower are adjusted by flow controllers LC1 and LC2;

[0045] (3) The temperature of the 46th tray is regulated by the temperature controller TC1;

[0046] (4) the purity of the overhead product ethylbenzene is regulated by regulating the overhead reflux flow;

[0047] (5) the purity of the product styrene at the bottom...

Embodiment 3

[0052] In this example, a THF / ethanol / water three-tower pressure-swing rectification process is constructed through the chemical engineering simulation software Aspen Plus and Aspen dynamic, and the flow chart is as attached to the description Figure 4 As shown, in order to establish a soft-sensing model to predict the composition of the bottom product, a soft-sensing model is established with the water purity of the bottom product of pressure swing rectification column 1 as the object. Regardless of the influence of the variables of the latter two towers on it, a total of 400 hours of dynamic data were simulated, and the sampling interval was 36s. Select 20 hours of it for testing, evaluate the performance of the model, and add random white noise to the data. According to the evaluation indexes predicted by six different algorithms, the determination coefficient values ​​of all models are above 0.9, which proves that the soft sensor model of the bottom components of the dist...

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 soft measurement method for on-line detection of components in a special rectification process, which comprises the following steps of: 1, selecting a soft measurement model input variable by adopting a principal component analysis random forest combined variable, 2, introducing a generalized robust loss function into a limit gradient algorithm, and optimizing a loss function hyper-parameter by adopting a Bayesian method, 3, inputting variables through a soft measurement model, and training in combination with a limit gradient algorithm to obtain a trained ARXGBoost model; 4, substituting real-time data into the ARXGBoost model for online detection; the adaptive updating robust limit gradient lifting algorithm has excellent prediction performance, realizes automatic updating of soft measurement, enhances the adaptive capability of the algorithm, improves the prediction reliability, performs online measurement and prediction on components, realizes direct control of the components, adopts adaptive adjustment robust parameters, and has the advantages of high robustness and high reliability. The prediction effect of the dynamic stage is good.

Description

technical field [0001] The invention relates to the technical field of rectification process control, in particular to a soft sensor method for on-line detection of components in a special rectification process. Background technique [0002] Due to the large number of controlled quantities and complex interactions between components in the special distillation process, component controllers must be used in many control strategies to meet product quality control requirements. However, online detection of product components in actual industrial processes is still There are many problems. Most of the existing measurement methods have many problems such as low accuracy, measurement lag, poor reliability, and high cost. The measurement results are difficult to use for component control. However, soft measurement technology uses other measurement information that is easy to obtain and realizes it through calculation. Estimating the measured variables, realizing the indirect measur...

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): G06F30/27G06K9/62G06Q10/04B01D3/14
CPCG06F30/27B01D3/14G06Q10/04G06F18/2135G06F18/214
Inventor 单宝明郭鲁钰杜康张方坤徐啟蕾朱兆友
Owner QINGDAO UNIV OF SCI & 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