Online reduced kernel PLS combined with GLRT for fault detection in chemical systems

R. Fazai, Majdi Mansouri, K. Abodayeh, Hazem Nounou, Mohamed Nounou

Research output: Contribution to journalReview article

Abstract

In this paper, an improved fault detection method is proposed based on kernel partial least squares (KPLS) model and generalized likelihood ratio test (GLRT) detection chart in order to enhance the monitoring abilities of nonlinear chemical processes. To deal with both high computational cost for large data set and time-varying dynamics of industrial processes, the proposed method is used to select a reduced set of kernel functions to build the KPLS model and applies it for online fault detection based on GLRT chart. Comparing with the conventional KPLS technique, the proposed method has the advantages of improving the computation efficiency by decreasing the dimension of kernel matrix. The developed online reduced KPLS-based GLRT (OR-GLRT) could improve the fault detection efficiency since it is able to track the time-varying characteristics of the processes. The fault detection performance of the developed OR-GLRT method is evaluated using a Tennessee Eastman process. The simulation results show that the proposed method outperforms the conventional GLRT technique.

Original languageEnglish
Pages (from-to)228-243
Number of pages16
JournalProcess Safety and Environmental Protection
Volume128
DOIs
Publication statusPublished - 1 Aug 2019

Fingerprint

Fault detection
detection method
chemical process
matrix
detection
method
chemical
test
Monitoring
monitoring
cost
simulation
Costs

Keywords

  • Chemical processes
  • Dynamic process
  • Fault detection
  • Generalized likelihood ratio test (GLRT)
  • Kernel PLS (KPLS)
  • Online reduced GLRT (OR-GLRT)

ASJC Scopus subject areas

  • Environmental Engineering
  • Environmental Chemistry
  • Chemical Engineering(all)
  • Safety, Risk, Reliability and Quality

Cite this

Online reduced kernel PLS combined with GLRT for fault detection in chemical systems. / Fazai, R.; Mansouri, Majdi; Abodayeh, K.; Nounou, Hazem; Nounou, Mohamed.

In: Process Safety and Environmental Protection, Vol. 128, 01.08.2019, p. 228-243.

Research output: Contribution to journalReview article

@article{806bdd98106a4f88bc4456f26429154a,
title = "Online reduced kernel PLS combined with GLRT for fault detection in chemical systems",
abstract = "In this paper, an improved fault detection method is proposed based on kernel partial least squares (KPLS) model and generalized likelihood ratio test (GLRT) detection chart in order to enhance the monitoring abilities of nonlinear chemical processes. To deal with both high computational cost for large data set and time-varying dynamics of industrial processes, the proposed method is used to select a reduced set of kernel functions to build the KPLS model and applies it for online fault detection based on GLRT chart. Comparing with the conventional KPLS technique, the proposed method has the advantages of improving the computation efficiency by decreasing the dimension of kernel matrix. The developed online reduced KPLS-based GLRT (OR-GLRT) could improve the fault detection efficiency since it is able to track the time-varying characteristics of the processes. The fault detection performance of the developed OR-GLRT method is evaluated using a Tennessee Eastman process. The simulation results show that the proposed method outperforms the conventional GLRT technique.",
keywords = "Chemical processes, Dynamic process, Fault detection, Generalized likelihood ratio test (GLRT), Kernel PLS (KPLS), Online reduced GLRT (OR-GLRT)",
author = "R. Fazai and Majdi Mansouri and K. Abodayeh and Hazem Nounou and Mohamed Nounou",
year = "2019",
month = "8",
day = "1",
doi = "10.1016/j.psep.2019.05.018",
language = "English",
volume = "128",
pages = "228--243",
journal = "Process Safety and Environmental Protection",
issn = "0957-5820",
publisher = "Institution of Chemical Engineers",

}

TY - JOUR

T1 - Online reduced kernel PLS combined with GLRT for fault detection in chemical systems

AU - Fazai, R.

AU - Mansouri, Majdi

AU - Abodayeh, K.

AU - Nounou, Hazem

AU - Nounou, Mohamed

PY - 2019/8/1

Y1 - 2019/8/1

N2 - In this paper, an improved fault detection method is proposed based on kernel partial least squares (KPLS) model and generalized likelihood ratio test (GLRT) detection chart in order to enhance the monitoring abilities of nonlinear chemical processes. To deal with both high computational cost for large data set and time-varying dynamics of industrial processes, the proposed method is used to select a reduced set of kernel functions to build the KPLS model and applies it for online fault detection based on GLRT chart. Comparing with the conventional KPLS technique, the proposed method has the advantages of improving the computation efficiency by decreasing the dimension of kernel matrix. The developed online reduced KPLS-based GLRT (OR-GLRT) could improve the fault detection efficiency since it is able to track the time-varying characteristics of the processes. The fault detection performance of the developed OR-GLRT method is evaluated using a Tennessee Eastman process. The simulation results show that the proposed method outperforms the conventional GLRT technique.

AB - In this paper, an improved fault detection method is proposed based on kernel partial least squares (KPLS) model and generalized likelihood ratio test (GLRT) detection chart in order to enhance the monitoring abilities of nonlinear chemical processes. To deal with both high computational cost for large data set and time-varying dynamics of industrial processes, the proposed method is used to select a reduced set of kernel functions to build the KPLS model and applies it for online fault detection based on GLRT chart. Comparing with the conventional KPLS technique, the proposed method has the advantages of improving the computation efficiency by decreasing the dimension of kernel matrix. The developed online reduced KPLS-based GLRT (OR-GLRT) could improve the fault detection efficiency since it is able to track the time-varying characteristics of the processes. The fault detection performance of the developed OR-GLRT method is evaluated using a Tennessee Eastman process. The simulation results show that the proposed method outperforms the conventional GLRT technique.

KW - Chemical processes

KW - Dynamic process

KW - Fault detection

KW - Generalized likelihood ratio test (GLRT)

KW - Kernel PLS (KPLS)

KW - Online reduced GLRT (OR-GLRT)

UR - http://www.scopus.com/inward/record.url?scp=85067174722&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85067174722&partnerID=8YFLogxK

U2 - 10.1016/j.psep.2019.05.018

DO - 10.1016/j.psep.2019.05.018

M3 - Review article

VL - 128

SP - 228

EP - 243

JO - Process Safety and Environmental Protection

JF - Process Safety and Environmental Protection

SN - 0957-5820

ER -