Feature Selection and Fault Classification of Reciprocating Compressors using a Genetic Algorithm and a Probabilistic Neural Network

M. Ahmed, F. Gu, A. Ball

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

Reciprocating compressors are widely used in industry for various purposes and faults occurring in them can degrade their performance, consume additional energy and even cause severe damage to the machine. Vibration monitoring techniques are often used for early fault detection and diagnosis, but it is difficult to prescribe a given set of effective diagnostic features because of the wide variety of operating conditions and the complexity of the vibration signals which originate from the many different vibrating and impact sources. This paper studies the use of genetic algorithms (GAs) and neural networks (NNs) to select effective diagnostic features for the fault diagnosis of a reciprocating compressor. A large number of common features are calculated from the time and frequency domains and envelope analysis. Applying GAs and NNs to these features found that envelope analysis has the most potential for differentiating three common faults: valve leakage, inter-cooler leakage and a loose drive belt. Simultaneously, the spread parameter of the probabilistic NN was also optimised. The selected subsets of features were examined based on vibration source characteristics. The approach developed and the trained NN are confirmed as possessing general characteristics for fault detection and diagnosis.

Original languageEnglish
Article number012112
JournalJournal of Physics: Conference Series
Volume305
Issue number1
DOIs
Publication statusPublished - 19 Jul 2011
Event9th International Conference on Damage Assessment of Structures - University of Oxford, Oxford, United Kingdom
Duration: 11 Jul 201113 Jul 2011
Conference number: 9
http://iopscience.iop.org/issue/1742-6596/305/1 (Link to Conference Proceedings)

Fingerprint

compressors
genetic algorithms
fault detection
vibration
leakage
envelopes
coolers
set theory
industries
damage
causes
energy

Cite this

@article{9cac199b38804f9cb08279e71670df89,
title = "Feature Selection and Fault Classification of Reciprocating Compressors using a Genetic Algorithm and a Probabilistic Neural Network",
abstract = "Reciprocating compressors are widely used in industry for various purposes and faults occurring in them can degrade their performance, consume additional energy and even cause severe damage to the machine. Vibration monitoring techniques are often used for early fault detection and diagnosis, but it is difficult to prescribe a given set of effective diagnostic features because of the wide variety of operating conditions and the complexity of the vibration signals which originate from the many different vibrating and impact sources. This paper studies the use of genetic algorithms (GAs) and neural networks (NNs) to select effective diagnostic features for the fault diagnosis of a reciprocating compressor. A large number of common features are calculated from the time and frequency domains and envelope analysis. Applying GAs and NNs to these features found that envelope analysis has the most potential for differentiating three common faults: valve leakage, inter-cooler leakage and a loose drive belt. Simultaneously, the spread parameter of the probabilistic NN was also optimised. The selected subsets of features were examined based on vibration source characteristics. The approach developed and the trained NN are confirmed as possessing general characteristics for fault detection and diagnosis.",
keywords = "Artificial Neural Networks, Condition Monitoring, Genetic algorithms and features extraction, Reciprocating Compressor",
author = "M. Ahmed and F. Gu and A. Ball",
year = "2011",
month = "7",
day = "19",
doi = "10.1088/1742-6596/305/1/012112",
language = "English",
volume = "305",
journal = "Journal of Physics: Conference Series",
issn = "1742-6588",
publisher = "IOP Publishing Ltd.",
number = "1",

}

TY - JOUR

T1 - Feature Selection and Fault Classification of Reciprocating Compressors using a Genetic Algorithm and a Probabilistic Neural Network

AU - Ahmed, M.

AU - Gu, F.

AU - Ball, A.

PY - 2011/7/19

Y1 - 2011/7/19

N2 - Reciprocating compressors are widely used in industry for various purposes and faults occurring in them can degrade their performance, consume additional energy and even cause severe damage to the machine. Vibration monitoring techniques are often used for early fault detection and diagnosis, but it is difficult to prescribe a given set of effective diagnostic features because of the wide variety of operating conditions and the complexity of the vibration signals which originate from the many different vibrating and impact sources. This paper studies the use of genetic algorithms (GAs) and neural networks (NNs) to select effective diagnostic features for the fault diagnosis of a reciprocating compressor. A large number of common features are calculated from the time and frequency domains and envelope analysis. Applying GAs and NNs to these features found that envelope analysis has the most potential for differentiating three common faults: valve leakage, inter-cooler leakage and a loose drive belt. Simultaneously, the spread parameter of the probabilistic NN was also optimised. The selected subsets of features were examined based on vibration source characteristics. The approach developed and the trained NN are confirmed as possessing general characteristics for fault detection and diagnosis.

AB - Reciprocating compressors are widely used in industry for various purposes and faults occurring in them can degrade their performance, consume additional energy and even cause severe damage to the machine. Vibration monitoring techniques are often used for early fault detection and diagnosis, but it is difficult to prescribe a given set of effective diagnostic features because of the wide variety of operating conditions and the complexity of the vibration signals which originate from the many different vibrating and impact sources. This paper studies the use of genetic algorithms (GAs) and neural networks (NNs) to select effective diagnostic features for the fault diagnosis of a reciprocating compressor. A large number of common features are calculated from the time and frequency domains and envelope analysis. Applying GAs and NNs to these features found that envelope analysis has the most potential for differentiating three common faults: valve leakage, inter-cooler leakage and a loose drive belt. Simultaneously, the spread parameter of the probabilistic NN was also optimised. The selected subsets of features were examined based on vibration source characteristics. The approach developed and the trained NN are confirmed as possessing general characteristics for fault detection and diagnosis.

KW - Artificial Neural Networks

KW - Condition Monitoring

KW - Genetic algorithms and features extraction

KW - Reciprocating Compressor

UR - http://www.scopus.com/inward/record.url?scp=80052052287&partnerID=8YFLogxK

U2 - 10.1088/1742-6596/305/1/012112

DO - 10.1088/1742-6596/305/1/012112

M3 - Article

VL - 305

JO - Journal of Physics: Conference Series

JF - Journal of Physics: Conference Series

SN - 1742-6588

IS - 1

M1 - 012112

ER -