Improved fault detection based on kernel PCA for monitoring industrial applications

Khadija Attouri; Majdi Mansouri; Mansour Hajji; Abdelmalek Kouadri; Abderrazak Bensmail; Kais Bouzrara; Hazem Nounou

doi:10.1016/j.jprocont.2023.103143

Improved fault detection based on kernel PCA for monitoring industrial applications

Khadija Attouri, Majdi Mansouri^*, Mansour Hajji, Abdelmalek Kouadri, Abderrazak Bensmail, Kais Bouzrara, Hazem Nounou

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

5 Citations (Scopus)

Abstract

The conventional Kernel Principal Component Analysis (KPCA) -based fault detection technique requires more computation time and memory storage space to analyze large-sized datasets. In this context, two techniques, Spectral Clustering (SpC) and Random Sampling (RnS), are developed to reduce the dataset size by retaining the more relevant observations while preserving the main statistical characteristics of the original dataset. These two techniques and others use the training dataset from two different industrial processes, Tennessee Eastman (TEP) and Cement Plant (CP) to be reduced and provided to build the Reduced KPCA (RKPCA) model-based fault detection scheme. The obtained results show the effectiveness of the proposed techniques in terms of some fault detection performance indices and computation costs.

Original language	English
Article number	103143
Number of pages	15
Journal	Journal of Process Control
Volume	133
DOIs	https://doi.org/10.1016/j.jprocont.2023.103143
Publication status	Published - Jan 2024
Externally published	Yes

Keywords

Cement Plant
Fault detection (FD)
Random Sampling (RnS)
Reduced Kernel Principal Component Analysis (RKPCA)
Spectral Clustering (SpC)
Tennessee Eastman process (TEP)

Access to Document

10.1016/j.jprocont.2023.103143

Cite this

@article{8d86769480e54879a94e33acf3b82acc,

title = "Improved fault detection based on kernel PCA for monitoring industrial applications",

abstract = "The conventional Kernel Principal Component Analysis (KPCA) -based fault detection technique requires more computation time and memory storage space to analyze large-sized datasets. In this context, two techniques, Spectral Clustering (SpC) and Random Sampling (RnS), are developed to reduce the dataset size by retaining the more relevant observations while preserving the main statistical characteristics of the original dataset. These two techniques and others use the training dataset from two different industrial processes, Tennessee Eastman (TEP) and Cement Plant (CP) to be reduced and provided to build the Reduced KPCA (RKPCA) model-based fault detection scheme. The obtained results show the effectiveness of the proposed techniques in terms of some fault detection performance indices and computation costs.",

keywords = "Cement Plant, Fault detection (FD), Random Sampling (RnS), Reduced Kernel Principal Component Analysis (RKPCA), Spectral Clustering (SpC), Tennessee Eastman process (TEP)",

author = "Khadija Attouri and Majdi Mansouri and Mansour Hajji and Abdelmalek Kouadri and Abderrazak Bensmail and Kais Bouzrara and Hazem Nounou",

note = "Publisher Copyright: {\textcopyright} 2023",

year = "2024",

month = jan,

doi = "10.1016/j.jprocont.2023.103143",

language = "English",

volume = "133",

journal = "Journal of Process Control",

issn = "0959-1524",

publisher = "Elsevier Ltd",

}

TY - JOUR

T1 - Improved fault detection based on kernel PCA for monitoring industrial applications

AU - Attouri, Khadija

AU - Mansouri, Majdi

AU - Hajji, Mansour

AU - Kouadri, Abdelmalek

AU - Bensmail, Abderrazak

AU - Bouzrara, Kais

AU - Nounou, Hazem

PY - 2024/1

Y1 - 2024/1

N2 - The conventional Kernel Principal Component Analysis (KPCA) -based fault detection technique requires more computation time and memory storage space to analyze large-sized datasets. In this context, two techniques, Spectral Clustering (SpC) and Random Sampling (RnS), are developed to reduce the dataset size by retaining the more relevant observations while preserving the main statistical characteristics of the original dataset. These two techniques and others use the training dataset from two different industrial processes, Tennessee Eastman (TEP) and Cement Plant (CP) to be reduced and provided to build the Reduced KPCA (RKPCA) model-based fault detection scheme. The obtained results show the effectiveness of the proposed techniques in terms of some fault detection performance indices and computation costs.

AB - The conventional Kernel Principal Component Analysis (KPCA) -based fault detection technique requires more computation time and memory storage space to analyze large-sized datasets. In this context, two techniques, Spectral Clustering (SpC) and Random Sampling (RnS), are developed to reduce the dataset size by retaining the more relevant observations while preserving the main statistical characteristics of the original dataset. These two techniques and others use the training dataset from two different industrial processes, Tennessee Eastman (TEP) and Cement Plant (CP) to be reduced and provided to build the Reduced KPCA (RKPCA) model-based fault detection scheme. The obtained results show the effectiveness of the proposed techniques in terms of some fault detection performance indices and computation costs.

KW - Cement Plant

KW - Fault detection (FD)

KW - Random Sampling (RnS)

KW - Reduced Kernel Principal Component Analysis (RKPCA)

KW - Spectral Clustering (SpC)

KW - Tennessee Eastman process (TEP)

UR - http://www.scopus.com/inward/record.url?scp=85179475292&partnerID=8YFLogxK

U2 - 10.1016/j.jprocont.2023.103143

DO - 10.1016/j.jprocont.2023.103143

M3 - Article

AN - SCOPUS:85179475292

SN - 0959-1524

VL - 133

JO - Journal of Process Control

JF - Journal of Process Control

M1 - 103143

ER -

Improved fault detection based on kernel PCA for monitoring industrial applications

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this