Life-cycle Modelling for Fault Detection - Extraction of PCA Models from Flowsheeting Simulators

Barbara Farsang, Sandor Nemeth, J. Abonyi

Research output: Contribution to journalArticle

Abstract

The operation of chemical processes is often supported by flowsheeting simulators and process monitoring systems. In many practical applications simulators used for planning, scheduling or operator training are often too complex for direct usage in realtime process monitoring; the structure of the related non-linear models does not support low-cost and rapid implementation of process monitoring systems. In this paper we present a novel method that effectively utilizes these first principle models for the development, maintenance and validation of multivariate statistical models. We demonstrate that the performance of Principal Component Analysis (PCA) models used for process monitoring can be improved by model based data reconciliation. The applicability of developed method is demonstrated in the Tennessee Eastman benchmark problem.

Original languageEnglish
Pages (from-to)421-426
Number of pages6
JournalComputer Aided Chemical Engineering
Volume33
DOIs
Publication statusPublished - 2014

Fingerprint

Process monitoring
Fault detection
Principal component analysis
Life cycle
Simulators
Scheduling
Planning
Costs

Keywords

  • Data reconciliation
  • Fault detection
  • Life-cycle modelling
  • PCA
  • Tennessee Eastman problem

ASJC Scopus subject areas

  • Chemical Engineering(all)
  • Computer Science Applications

Cite this

Life-cycle Modelling for Fault Detection - Extraction of PCA Models from Flowsheeting Simulators. / Farsang, Barbara; Nemeth, Sandor; Abonyi, J.

In: Computer Aided Chemical Engineering, Vol. 33, 2014, p. 421-426.

Research output: Contribution to journalArticle

@article{0cb03f2744bf46e6a0652459feb3fa50,
title = "Life-cycle Modelling for Fault Detection - Extraction of PCA Models from Flowsheeting Simulators",
abstract = "The operation of chemical processes is often supported by flowsheeting simulators and process monitoring systems. In many practical applications simulators used for planning, scheduling or operator training are often too complex for direct usage in realtime process monitoring; the structure of the related non-linear models does not support low-cost and rapid implementation of process monitoring systems. In this paper we present a novel method that effectively utilizes these first principle models for the development, maintenance and validation of multivariate statistical models. We demonstrate that the performance of Principal Component Analysis (PCA) models used for process monitoring can be improved by model based data reconciliation. The applicability of developed method is demonstrated in the Tennessee Eastman benchmark problem.",
keywords = "Data reconciliation, Fault detection, Life-cycle modelling, PCA, Tennessee Eastman problem",
author = "Barbara Farsang and Sandor Nemeth and J. Abonyi",
year = "2014",
doi = "10.1016/B978-0-444-63456-6.50071-5",
language = "English",
volume = "33",
pages = "421--426",
journal = "Computer Aided Chemical Engineering",
issn = "1570-7946",
publisher = "Elsevier",

}

TY - JOUR

T1 - Life-cycle Modelling for Fault Detection - Extraction of PCA Models from Flowsheeting Simulators

AU - Farsang, Barbara

AU - Nemeth, Sandor

AU - Abonyi, J.

PY - 2014

Y1 - 2014

N2 - The operation of chemical processes is often supported by flowsheeting simulators and process monitoring systems. In many practical applications simulators used for planning, scheduling or operator training are often too complex for direct usage in realtime process monitoring; the structure of the related non-linear models does not support low-cost and rapid implementation of process monitoring systems. In this paper we present a novel method that effectively utilizes these first principle models for the development, maintenance and validation of multivariate statistical models. We demonstrate that the performance of Principal Component Analysis (PCA) models used for process monitoring can be improved by model based data reconciliation. The applicability of developed method is demonstrated in the Tennessee Eastman benchmark problem.

AB - The operation of chemical processes is often supported by flowsheeting simulators and process monitoring systems. In many practical applications simulators used for planning, scheduling or operator training are often too complex for direct usage in realtime process monitoring; the structure of the related non-linear models does not support low-cost and rapid implementation of process monitoring systems. In this paper we present a novel method that effectively utilizes these first principle models for the development, maintenance and validation of multivariate statistical models. We demonstrate that the performance of Principal Component Analysis (PCA) models used for process monitoring can be improved by model based data reconciliation. The applicability of developed method is demonstrated in the Tennessee Eastman benchmark problem.

KW - Data reconciliation

KW - Fault detection

KW - Life-cycle modelling

KW - PCA

KW - Tennessee Eastman problem

UR - http://www.scopus.com/inward/record.url?scp=84902950717&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84902950717&partnerID=8YFLogxK

U2 - 10.1016/B978-0-444-63456-6.50071-5

DO - 10.1016/B978-0-444-63456-6.50071-5

M3 - Article

VL - 33

SP - 421

EP - 426

JO - Computer Aided Chemical Engineering

JF - Computer Aided Chemical Engineering

SN - 1570-7946

ER -