Publication detail

Comparison of Performance of Machine Learning Methods for Bearing Faults Classification Using Time-Domain Features

DOSEDĚL, M. HAVRÁNEK, Z.

Original Title

Comparison of Performance of Machine Learning Methods for Bearing Faults Classification Using Time-Domain Features

English Title

Comparison of Performance of Machine Learning Methods for Bearing Faults Classification Using Time-Domain Features

Type

conference paper

Language

en

Original Abstract

This article deals with comparison of a classification success rate of machine learning methods used for vibration signals captured on damaged bearing. The most significant and the most used methods suitable for this vibrodiagnostic field have been selected, namely support vector machine, k-nearest neighbor, Mahalanobis-Taguchi system and principal component analysis for reduction of a dimensionality of a features space. All the methods have been implemented in MATLAB and their performance has been analyzed using same input datasets. As the input for all the aforementioned methods, data from bearing shortened life-cycle test, done by Bearing Data Center at Case Western Reserve University, have been used. Intentionally, computationally intensive pre-processing procedures have been excluded from the computational chain, as only time domain features have been used for the classification process. Resulting from this study, a classification error rate, an execution time, an implementation difficulty and a robustness of the algorithm strongly differ among all the methods while using the same input data and the same number and type of the input extracted features.

English abstract

This article deals with comparison of a classification success rate of machine learning methods used for vibration signals captured on damaged bearing. The most significant and the most used methods suitable for this vibrodiagnostic field have been selected, namely support vector machine, k-nearest neighbor, Mahalanobis-Taguchi system and principal component analysis for reduction of a dimensionality of a features space. All the methods have been implemented in MATLAB and their performance has been analyzed using same input datasets. As the input for all the aforementioned methods, data from bearing shortened life-cycle test, done by Bearing Data Center at Case Western Reserve University, have been used. Intentionally, computationally intensive pre-processing procedures have been excluded from the computational chain, as only time domain features have been used for the classification process. Resulting from this study, a classification error rate, an execution time, an implementation difficulty and a robustness of the algorithm strongly differ among all the methods while using the same input data and the same number and type of the input extracted features.

Keywords

machine learning, support vector machine, k-nearest neighbor, Mahalanobis-Taguchi system, principal component analysis, vibrodiagnostics

Released

24.11.2020

ISBN

978-1-7281-5600-2

Book

Proceedings of the 2020 19th International Conference on Mechatronics – Mechatronika (ME)

Edition

1st edition

Pages from

140

Pages to

146

Pages count

341

URL

Documents

BibTex


@inproceedings{BUT165683,
  author="Martin {Doseděl} and Zdeněk {Havránek}",
  title="Comparison of Performance of Machine Learning Methods for Bearing Faults Classification Using Time-Domain Features",
  annote="This article deals with comparison of a classification success rate of machine learning methods used for vibration signals captured on damaged bearing. The most significant and the most used methods suitable for this vibrodiagnostic field have been selected, namely support vector machine, k-nearest neighbor, Mahalanobis-Taguchi system and principal component analysis for reduction of a dimensionality of a features space. All the methods have been implemented in MATLAB and their performance has been analyzed using same input datasets. As the input for all the aforementioned methods, data from bearing shortened life-cycle test, done by Bearing Data Center at Case Western Reserve University, have been used. Intentionally, computationally intensive pre-processing procedures have been excluded from the computational chain, as only time domain features have been used for the classification process.  Resulting from this study, a classification error rate, an execution time, an implementation difficulty and a robustness of the algorithm strongly differ among all the methods while using the same input data and the same number and type of the input extracted features.",
  booktitle="Proceedings of the 2020 19th International Conference on Mechatronics – Mechatronika (ME)",
  chapter="165683",
  edition="1st edition",
  howpublished="online",
  year="2020",
  month="november",
  pages="140--146",
  type="conference paper"
}