Data driven approach to monitoring and fault detection in process control plants

Linda Varghese, V. I. George, Krishnamoorthi Makkithaya, Abhishek Kumar

Research output: Contribution to journalArticle

Abstract

Detecting faults and monitoring is a very important activity in process control plants for increasing the efficiency of the plant. Data driven approach is particularly helpful for fault detection, if the underlying mathematical model of the process is very complex. It can be used to create automatic systems which accurately predict whether the operating condition of the plant is normal or faulty. This paper compares different supervised learning algorithms, in order to detect fault in process control plant. The algorithms are tested in Matlab environment. Finally, all the models give satisfactory accuracy while detecting two different types of faults as well as normal operating condition.

Original languageEnglish
Pages (from-to)1121-1128
Number of pages8
JournalInternational Journal of Control Theory and Applications
Volume8
Issue number3
Publication statusPublished - 2015

Fingerprint

Fault detection
Process control
Monitoring
Supervised learning
Learning algorithms
Mathematical models

All Science Journal Classification (ASJC) codes

  • Computer Science(all)

Cite this

@article{1f4337412ee349689f305bef38533575,
title = "Data driven approach to monitoring and fault detection in process control plants",
abstract = "Detecting faults and monitoring is a very important activity in process control plants for increasing the efficiency of the plant. Data driven approach is particularly helpful for fault detection, if the underlying mathematical model of the process is very complex. It can be used to create automatic systems which accurately predict whether the operating condition of the plant is normal or faulty. This paper compares different supervised learning algorithms, in order to detect fault in process control plant. The algorithms are tested in Matlab environment. Finally, all the models give satisfactory accuracy while detecting two different types of faults as well as normal operating condition.",
author = "Linda Varghese and George, {V. I.} and Krishnamoorthi Makkithaya and Abhishek Kumar",
year = "2015",
language = "English",
volume = "8",
pages = "1121--1128",
journal = "International Journal of Control Theory and Applications",
issn = "0974-5572",
publisher = "Serials Publications",
number = "3",

}

Data driven approach to monitoring and fault detection in process control plants. / Varghese, Linda; George, V. I.; Makkithaya, Krishnamoorthi; Kumar, Abhishek.

In: International Journal of Control Theory and Applications, Vol. 8, No. 3, 2015, p. 1121-1128.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Data driven approach to monitoring and fault detection in process control plants

AU - Varghese, Linda

AU - George, V. I.

AU - Makkithaya, Krishnamoorthi

AU - Kumar, Abhishek

PY - 2015

Y1 - 2015

N2 - Detecting faults and monitoring is a very important activity in process control plants for increasing the efficiency of the plant. Data driven approach is particularly helpful for fault detection, if the underlying mathematical model of the process is very complex. It can be used to create automatic systems which accurately predict whether the operating condition of the plant is normal or faulty. This paper compares different supervised learning algorithms, in order to detect fault in process control plant. The algorithms are tested in Matlab environment. Finally, all the models give satisfactory accuracy while detecting two different types of faults as well as normal operating condition.

AB - Detecting faults and monitoring is a very important activity in process control plants for increasing the efficiency of the plant. Data driven approach is particularly helpful for fault detection, if the underlying mathematical model of the process is very complex. It can be used to create automatic systems which accurately predict whether the operating condition of the plant is normal or faulty. This paper compares different supervised learning algorithms, in order to detect fault in process control plant. The algorithms are tested in Matlab environment. Finally, all the models give satisfactory accuracy while detecting two different types of faults as well as normal operating condition.

UR - http://www.scopus.com/inward/record.url?scp=84958959080&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84958959080&partnerID=8YFLogxK

M3 - Article

VL - 8

SP - 1121

EP - 1128

JO - International Journal of Control Theory and Applications

JF - International Journal of Control Theory and Applications

SN - 0974-5572

IS - 3

ER -