Fault classification of reciprocating compressor based on Neural Networks and Support Vector Machines

M. Ahmed, S. Abdusslam, M. Baqqar, F. Gu, A. D. Ball

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

12 Citations (Scopus)

Abstract

Reciprocating compressors play a major part in many industrial systems and faults occurring in them can degrade performance, consume additional energy, cause severe damage to the machine and possibly even system shut-down. Traditional vibration monitoring techniques have found it difficult to determine a set of effective diagnostic features due to the high complexity of the vibration signals because of the many different impact sources and wide range of practical operating conditions. This paper focuses on the development of an advanced signal classifier for a reciprocating compressor using vibration signals. Artificial Neural Networks (ANN) and Support Vector Machines (SVM) have been applied, trained and tested for feature extraction and fault classification. The accuracy of both techniques is compared to determine the optimum fault classifier. The results show that the model behaves well, and classification rate accuracy is up to 100% for both binary classes (a single fault present in the compressor) and multi-classes (three faults present).

Original languageEnglish
Title of host publication17th International Conference on Automation and Computing, ICAC 2011
PublisherIEEE
Pages213-218
Number of pages6
ISBN (Electronic)9781862180987
ISBN (Print)9781467300001
Publication statusPublished - 21 Nov 2011
Event17th International Conference on Automation and Computing - Huddersfield, United Kingdom
Duration: 10 Sep 201110 Sep 2011
Conference number: 17
http://www.wikicfp.com/cfp/servlet/event.showcfp?eventid=14139&copyownerid=20561 (Link to Event Details)
https://www.tib.eu/en/search/id/TIBKAT%3A73636952X/

Conference

Conference17th International Conference on Automation and Computing
Abbreviated titleICAC 2011
Country/TerritoryUnited Kingdom
CityHuddersfield
Period10/09/1110/09/11
Internet address

Fingerprint

Dive into the research topics of 'Fault classification of reciprocating compressor based on Neural Networks and Support Vector Machines'. Together they form a unique fingerprint.

Cite this