On-line detection of homogeneous operation ranges by dynamic principal component analysis based time-series segmentation

László Dobos, J. Abonyi

Research output: Contribution to journalArticle

23 Citations (Scopus)

Abstract

Development of chemical process technologies shall be based on the analysis of process data. In the field of process monitoring the recursive Principal Component Analysis (PCA) is widely applied to detect any misbehavior of the technology. The investigation of transient states needs dynamic PCA to describe the dynamic behavior more accurately. By combining and integrating the recursive and dynamic PCA into time series segmentation techniques, efficient multivariate segmentation methods were resulted to detect homogenous operation ranges based on process data. The similarity of time-series segments is evaluated based on the Krzanowski-similarity factor, which compares the hyperplanes determined by the PCA models. With the help of developed time series segmentation framework, separation of operation regimes becomes possible for supporting process monitoring and control. The performance of the proposed methodology is presented throughout a linear process and the commonly applied Tennessee Eastman process.

Original languageEnglish
Pages (from-to)96-105
Number of pages10
JournalChemical Engineering Science
Volume75
DOIs
Publication statusPublished - Jun 18 2012

Fingerprint

Dynamic Analysis
Principal component analysis
Principal Component Analysis
Time series
Segmentation
Process Monitoring
Process monitoring
Range of data
Linear Process
Transient State
Chemical Processes
Process Control
Hyperplane
Dynamic Behavior
Process control
Methodology
Similarity
Model

Keywords

  • Data mining
  • Dynamic principlal component analysis
  • Process monitoring
  • Tennessee Eastman process
  • Time-series segmentation
  • Variable forgetting factor

ASJC Scopus subject areas

  • Chemical Engineering(all)
  • Chemistry(all)
  • Applied Mathematics
  • Industrial and Manufacturing Engineering

Cite this

On-line detection of homogeneous operation ranges by dynamic principal component analysis based time-series segmentation. / Dobos, László; Abonyi, J.

In: Chemical Engineering Science, Vol. 75, 18.06.2012, p. 96-105.

Research output: Contribution to journalArticle

@article{23fc8d48bf4742bcb6da2b509d8a97a3,
title = "On-line detection of homogeneous operation ranges by dynamic principal component analysis based time-series segmentation",
abstract = "Development of chemical process technologies shall be based on the analysis of process data. In the field of process monitoring the recursive Principal Component Analysis (PCA) is widely applied to detect any misbehavior of the technology. The investigation of transient states needs dynamic PCA to describe the dynamic behavior more accurately. By combining and integrating the recursive and dynamic PCA into time series segmentation techniques, efficient multivariate segmentation methods were resulted to detect homogenous operation ranges based on process data. The similarity of time-series segments is evaluated based on the Krzanowski-similarity factor, which compares the hyperplanes determined by the PCA models. With the help of developed time series segmentation framework, separation of operation regimes becomes possible for supporting process monitoring and control. The performance of the proposed methodology is presented throughout a linear process and the commonly applied Tennessee Eastman process.",
keywords = "Data mining, Dynamic principlal component analysis, Process monitoring, Tennessee Eastman process, Time-series segmentation, Variable forgetting factor",
author = "L{\'a}szl{\'o} Dobos and J. Abonyi",
year = "2012",
month = "6",
day = "18",
doi = "10.1016/j.ces.2012.02.022",
language = "English",
volume = "75",
pages = "96--105",
journal = "Chemical Engineering Science",
issn = "0009-2509",
publisher = "Elsevier BV",

}

TY - JOUR

T1 - On-line detection of homogeneous operation ranges by dynamic principal component analysis based time-series segmentation

AU - Dobos, László

AU - Abonyi, J.

PY - 2012/6/18

Y1 - 2012/6/18

N2 - Development of chemical process technologies shall be based on the analysis of process data. In the field of process monitoring the recursive Principal Component Analysis (PCA) is widely applied to detect any misbehavior of the technology. The investigation of transient states needs dynamic PCA to describe the dynamic behavior more accurately. By combining and integrating the recursive and dynamic PCA into time series segmentation techniques, efficient multivariate segmentation methods were resulted to detect homogenous operation ranges based on process data. The similarity of time-series segments is evaluated based on the Krzanowski-similarity factor, which compares the hyperplanes determined by the PCA models. With the help of developed time series segmentation framework, separation of operation regimes becomes possible for supporting process monitoring and control. The performance of the proposed methodology is presented throughout a linear process and the commonly applied Tennessee Eastman process.

AB - Development of chemical process technologies shall be based on the analysis of process data. In the field of process monitoring the recursive Principal Component Analysis (PCA) is widely applied to detect any misbehavior of the technology. The investigation of transient states needs dynamic PCA to describe the dynamic behavior more accurately. By combining and integrating the recursive and dynamic PCA into time series segmentation techniques, efficient multivariate segmentation methods were resulted to detect homogenous operation ranges based on process data. The similarity of time-series segments is evaluated based on the Krzanowski-similarity factor, which compares the hyperplanes determined by the PCA models. With the help of developed time series segmentation framework, separation of operation regimes becomes possible for supporting process monitoring and control. The performance of the proposed methodology is presented throughout a linear process and the commonly applied Tennessee Eastman process.

KW - Data mining

KW - Dynamic principlal component analysis

KW - Process monitoring

KW - Tennessee Eastman process

KW - Time-series segmentation

KW - Variable forgetting factor

UR - http://www.scopus.com/inward/record.url?scp=84860524308&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84860524308&partnerID=8YFLogxK

U2 - 10.1016/j.ces.2012.02.022

DO - 10.1016/j.ces.2012.02.022

M3 - Article

AN - SCOPUS:84860524308

VL - 75

SP - 96

EP - 105

JO - Chemical Engineering Science

JF - Chemical Engineering Science

SN - 0009-2509

ER -