Multi-period intermittent process soft measurement modeling method based on CSJITL-RVM
A modeling method and soft-sensing technology, applied in prediction, character and pattern recognition, data processing applications, etc., can solve the problem of reducing the online prediction accuracy of quality variables in the soft-sensing model, the low accuracy of similar label data selection results, and ignoring the process Data multi-period characteristics and timing characteristics and other issues
Pending Publication Date: 2020-04-28
BEIJING UNIV OF CHEM TECH
View PDF5 Cites 7 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, due to the multi-period, nonlinear characteristics of the intermittent process, and the time-series constraints of the process data, the soft sensor method based on JITL-RVM only considers the algebraic spatial similarity of the process data when predicting the label value of the unlabeled data. The multi-period characteristics and timing characteristics of process data are ignored, resulting in low accuracy of similar label data selection results for unlabeled data, directly affecting the label prediction results of unlabeled data, and reducing the online prediction accuracy of quality variables in the soft sensor model
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0059] Penicillin is an antibiotic with extensive clinical value, and its production process is a typical non-linear, dynamic and multi-period batch production process. Using the penicillin fermentation process simulation platform (PenSim v2.0) to generate 25 batches of training data and 4 batches of test data, the sampling time and sampling interval of each batch are 400h and 1h, and the sampling ratio of labeled data and unlabeled data is 1:9 . In the experiment, 11 process variables were selected for the soft-sensing modeling of the penicillin fermentation process, as shown in Table 1, in which the process variables with serial numbers 1-10 are auxiliary variables, and the process variables with serial number 11 are quality variables.
[0060] Table 1 Penicillin Fermentation Process Variables
[0061]
[0062]
[0063] The concrete steps that the present invention is applied to the penicillin fermentation process are as follows:
[0064] Step 1: The collected proces...
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
Login to View More Abstract
The invention discloses a multi-period intermittent process soft measurement modeling method based on CSJITL-RVM. The method comprises the steps of firstly, performing period division on a multi-period intermittent process by utilizing an SCFCM clustering method; then, introducing a label prediction method based on CSJITL; calculating the similarity between the label data and the label-free data by adopting a similarity factor integrating process data space similarity, time period similarity and time sequence similarity, obtaining a training data set of a label prediction model by screening similarity values, establishing the label prediction model, and realizing label prediction of the label-free data; and finally, fusing the label data and the predicted label-free data, and establishinga time period soft measurement model of the RVM to realize prediction of the batch process quality variable. According to the method, when the label value of the label-free data is predicted by usingthe label data, the space, time period and time sequence similarity of the process data is considered, the accuracy of the label prediction value is improved, accurate and effective training data areprovided for establishing an intermittent process soft measurement model, and the prediction precision of the quality variable is improved.
Description
technical field [0001] The invention belongs to the technical field of intermittent process soft sensing, and in particular relates to a multi-period intermittent process soft sensing modeling based on Comprehensive Similarity Just-in-time Learning-Relevant Vector Machine (CSJITL-RVM) method. Background technique [0002] As one of the main production methods of modern production, batch process has been widely used in chemical industry, food, medicine, semiconductor processing and other fields. To ensure its efficient, reliable and safe operation, online measurement of quality variables is required. Soft sensor technology is an online estimation technology for unmeasurable variables and difficult-to-measure variables. It realizes online prediction of quality variables by establishing a mathematical model between auxiliary variables and quality variables. It has been widely used in online measurement of quality variables in batch processes. . [0003] In the intermittent pr...
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
Login to View More IPC IPC(8): G06K9/62G06Q10/04
CPCG06Q10/04G06F18/23G06F18/22G06F18/214Y02P90/02
Inventor王建林邱科鹏潘佳
OwnerBEIJING UNIV OF CHEM TECH