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

Fermentation process stage division method based on dynamic feature extraction

A technology of dynamic characteristics and fermentation process, applied in the direction of instruments, adaptive control, control/regulation systems, etc., can solve the problems of inaccurate division of fermentation process stages and division of production process into multiple sub-stages, etc., to achieve the effect of improving accuracy

Active Publication Date: 2021-06-08
BEIJING UNIV OF TECH
View PDF12 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The multi-stage characteristics of the fermentation process are largely reflected by the dynamics of the characterization process changes. If the dynamic characteristics of the process are ignored, the entire production process cannot be accurately divided into multiple sub-stages. Therefore, the present invention adopts a design based on dynamic feature extraction. Method design based on ideas to solve the problem of inaccurate division of fermentation process stages in existing technologies

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
  • Fermentation process stage division method based on dynamic feature extraction
  • Fermentation process stage division method based on dynamic feature extraction
  • Fermentation process stage division method based on dynamic feature extraction

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0036] The Pensim simulation platform is an internationally influential penicillin simulation platform, and relevant research has shown its practicability and effectiveness. In this paper, based on this platform, the experimental simulation of the fermentation process of penicillin was carried out. The fermentation time of each batch was set to 400h, the sampling interval was 1h, and 11 process variables were selected (including ventilation rate, stirring power, substrate flow acceleration rate, substrate Temperature, dissolved oxygen concentration, exhaust carbon dioxide concentration, pH, temperature, reaction heat, cold water flow acceleration rate and substrate concentration) and 1 quality variable (product concentration) were monitored, and 40 batches of penicillin fermentation processes were selected for training in this paper Modeling, 20 batches for prediction.

[0037] Table 1 Penicillin Fermentation Process Variables

[0038] Table 1 Process variables of Penicillin ...

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 stage division method based on dynamic feature extraction, which is used for solving the problem that dynamic feature changes in the capturing process of an existing method are not sensitive enough. The stage division method comprises the following steps of: firstly, expanding original data along batches, performing partial least square analysis on each time slice matrix to obtain score matrixes of process variables and quality variables, and clustering the joint score matrixes by adopting an AP algorithm to realize first-step division; and extracting dynamic features representing process dynamics by adopting an encoder-decoder model, carrying out second-step division on the dynamic features by adopting the AP algorithm, comprehensively analyzing two-step division results, and finally dividing the production process into different stable stages and transition stages. After stage division is completed, an attention LSTM quality prediction model can be further established for each divided stage, and expansion application is carried out.

Description

technical field [0001] The invention relates to the technical field of data-driven process monitoring, in particular to a fermentation process stage division method based on dynamic feature extraction proposed for the multi-stage characteristics of the fermentation process. The present invention uses the joint scoring matrix to realize the first step of division; then uses the encoder-decoder model to extract the dynamic features representing the dynamics of the process to realize the second step of division, and comprehensively analyzes the results of the two-step division to finally divide the production process into different stable stages and transitional stages. Background technique [0002] As an extremely important production method in modern industrial processes, batch process has been widely used in the production of various and high value-added products such as medicine, food, biochemical industry, and semiconductor. However, in some production processes such as f...

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
CPCG05B13/042Y02P90/30
Inventor 高学金孟令军高慧慧
Owner BEIJING UNIV OF TECH
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