A new fault detection method for nonlinear process monitoring

Radhia Fazai, Okba Taouali, Mohamed-Faouzi Harkat, Nasereddine Bouguila

Research output: Contribution to journalArticle

23 Citations (Scopus)

Abstract

Kernel Principal Component Analysis (KPCA) is a nonlinear extension of Principal Component Analysis (PCA). Recently, it is the most popular technique for monitoring nonlinear processes. However, the time-varying property of the industrial processes requires the adaptive ability of the KPCA. Therefore, in this paper, a Variable Moving Window Kernel PCA (VMWKPCA) method is proposed to update the KPCA model. The concept of this method consists of varying the size of the moving window according to the change of the normal process. To evaluate the performance of the proposed method, the VMWKPCA is applied for monitoring a Continuous Stirred Tank Reactor (CSTR) and a Tennessee Eastman process (TE). The results are satisfactory compared to the conventional Moving Window Kernel PCA (MWKPCA) and the Adaptive Kernel PCA (AKPCA).

Original languageEnglish
Pages (from-to)3425-3436
Number of pages12
JournalInternational Journal of Advanced Manufacturing Technology
Volume87
Issue number9-12
DOIs
Publication statusPublished - 1 Dec 2016
Externally publishedYes

Fingerprint

Process monitoring
Fault detection
Principal component analysis
Monitoring

Keywords

  • AKPCA
  • Fault detection
  • KPCA
  • MWKPCA
  • Variable Moving Window Kernel PCA (VMWKPCA)

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Software
  • Mechanical Engineering
  • Computer Science Applications
  • Industrial and Manufacturing Engineering

Cite this

A new fault detection method for nonlinear process monitoring. / Fazai, Radhia; Taouali, Okba; Harkat, Mohamed-Faouzi; Bouguila, Nasereddine.

In: International Journal of Advanced Manufacturing Technology, Vol. 87, No. 9-12, 01.12.2016, p. 3425-3436.

Research output: Contribution to journalArticle

Fazai, Radhia ; Taouali, Okba ; Harkat, Mohamed-Faouzi ; Bouguila, Nasereddine. / A new fault detection method for nonlinear process monitoring. In: International Journal of Advanced Manufacturing Technology. 2016 ; Vol. 87, No. 9-12. pp. 3425-3436.
@article{7edeac065ce146a6937c822cd4ea2408,
title = "A new fault detection method for nonlinear process monitoring",
abstract = "Kernel Principal Component Analysis (KPCA) is a nonlinear extension of Principal Component Analysis (PCA). Recently, it is the most popular technique for monitoring nonlinear processes. However, the time-varying property of the industrial processes requires the adaptive ability of the KPCA. Therefore, in this paper, a Variable Moving Window Kernel PCA (VMWKPCA) method is proposed to update the KPCA model. The concept of this method consists of varying the size of the moving window according to the change of the normal process. To evaluate the performance of the proposed method, the VMWKPCA is applied for monitoring a Continuous Stirred Tank Reactor (CSTR) and a Tennessee Eastman process (TE). The results are satisfactory compared to the conventional Moving Window Kernel PCA (MWKPCA) and the Adaptive Kernel PCA (AKPCA).",
keywords = "AKPCA, Fault detection, KPCA, MWKPCA, Variable Moving Window Kernel PCA (VMWKPCA)",
author = "Radhia Fazai and Okba Taouali and Mohamed-Faouzi Harkat and Nasereddine Bouguila",
year = "2016",
month = "12",
day = "1",
doi = "10.1007/s00170-016-8745-7",
language = "English",
volume = "87",
pages = "3425--3436",
journal = "International Journal of Advanced Manufacturing Technology",
issn = "0268-3768",
publisher = "Springer London",
number = "9-12",

}

TY - JOUR

T1 - A new fault detection method for nonlinear process monitoring

AU - Fazai, Radhia

AU - Taouali, Okba

AU - Harkat, Mohamed-Faouzi

AU - Bouguila, Nasereddine

PY - 2016/12/1

Y1 - 2016/12/1

N2 - Kernel Principal Component Analysis (KPCA) is a nonlinear extension of Principal Component Analysis (PCA). Recently, it is the most popular technique for monitoring nonlinear processes. However, the time-varying property of the industrial processes requires the adaptive ability of the KPCA. Therefore, in this paper, a Variable Moving Window Kernel PCA (VMWKPCA) method is proposed to update the KPCA model. The concept of this method consists of varying the size of the moving window according to the change of the normal process. To evaluate the performance of the proposed method, the VMWKPCA is applied for monitoring a Continuous Stirred Tank Reactor (CSTR) and a Tennessee Eastman process (TE). The results are satisfactory compared to the conventional Moving Window Kernel PCA (MWKPCA) and the Adaptive Kernel PCA (AKPCA).

AB - Kernel Principal Component Analysis (KPCA) is a nonlinear extension of Principal Component Analysis (PCA). Recently, it is the most popular technique for monitoring nonlinear processes. However, the time-varying property of the industrial processes requires the adaptive ability of the KPCA. Therefore, in this paper, a Variable Moving Window Kernel PCA (VMWKPCA) method is proposed to update the KPCA model. The concept of this method consists of varying the size of the moving window according to the change of the normal process. To evaluate the performance of the proposed method, the VMWKPCA is applied for monitoring a Continuous Stirred Tank Reactor (CSTR) and a Tennessee Eastman process (TE). The results are satisfactory compared to the conventional Moving Window Kernel PCA (MWKPCA) and the Adaptive Kernel PCA (AKPCA).

KW - AKPCA

KW - Fault detection

KW - KPCA

KW - MWKPCA

KW - Variable Moving Window Kernel PCA (VMWKPCA)

UR - http://www.scopus.com/inward/record.url?scp=84963755478&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84963755478&partnerID=8YFLogxK

U2 - 10.1007/s00170-016-8745-7

DO - 10.1007/s00170-016-8745-7

M3 - Article

VL - 87

SP - 3425

EP - 3436

JO - International Journal of Advanced Manufacturing Technology

JF - International Journal of Advanced Manufacturing Technology

SN - 0268-3768

IS - 9-12

ER -