CoChat

Effectiveness Analysis of PMSM Motor Rolling Bearing Fault Detectors Based on Vibration Analysis and Shallow Neural Networks

Open access
Ewert, Pawel, Orlowska-Kowalska, Teresa, Jankowska, Kamila · 2021 · Energies

The DOI matches its Crossref record; 48 of the 48 references it cites were found in a registry or an index.

Abstract

Permanent magnet synchronous motors (PMSMs) are becoming more popular, both in industrial applications and in electric and hybrid vehicle drives. Unfortunately, like the others, these are not reliable drives. As in the drive systems with induction motors, the rolling bearings can often fail. This paper focuses on the possibility of detecting this type of mechanical damage by analysing mechanical vibrations supported by shallow neural networks (NNs). For the extraction of diagnostic symptoms, the Fast Fourier Transform (FFT) and the Hilbert transform (HT) were used to obtain the envelope signal, which was subjected to the FFT analysis. Three types of neural networks were tested to automate the detection process: multilayer perceptron (MLP), neural network with radial base function (RBF), and Kohonen map (self-organizing map, SOM). The input signals of these networks were the amplitudes of harmonic components characteristic of damage to bearing elements, obtained as a result of FFT or HT analysis of the vibration acceleration signal. The effectiveness of the analysed NN structures was compared from the point of view of the influence of the network architecture and various parameters of the learning process on the detection effectiveness.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.3390/en14030712

Cites

48 of 48 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Diagram of the construction of a rolling bearing. Figure 1. Diagram of the construction of a rolling bearing.
Figure 1 Diagram of the construction of a rolling bearing. Figure 1. Diagram of the construction of a rolling bearing.CC-BY · From the PDF
Figure 2 Example of structure of the feedforward multilayer perceptron (MLP) network. Figure 2. Example of structure of the feedforward multilayer perceptron (MLP) network.
Figure 2 Example of structure of the feedforward multilayer perceptron (MLP) network. Figure 2. Example of structure of the feedforward multilayer perceptron (MLP) network.CC-BY · From the PDF
Figure 3 General structures of radial basis function (RBF) network. Figure 3. General structures of radial basis function (RBF) network.
Figure 3 General structures of radial basis function (RBF) network. Figure 3. General structures of radial basis function (RBF) network.CC-BY · From the PDF
Figure 4 General architecture of Kohonen map. Figure 4. General architecture of Kohonen map.
Figure 4 General architecture of Kohonen map. Figure 4. General architecture of Kohonen map.CC-BY · From the PDF
Figure 5 Schematic diagram and photo of the laboratory stand. Figure 5. Schematic diagram and photo of the laboratory stand.
Figure 5 Schematic diagram and photo of the laboratory stand. Figure 5. Schematic diagram and photo of the laboratory stand.CC-BY · From the PDF
Image 6
CC-BY · From the PDF
Figure 7 Examples of harmonic amplitudes with characteristic frequencies obtained using envelope (ENV) for vibration acceleration signal; different load torques and supplying frequencies of Figure 7. Examples of harmonic amplitudes with characteristic frequencies obtained using envelope (ENV) for vibration acceleration signal; different load torques and supplying frequencies of PMSM drive.
Figure 7 Examples of harmonic amplitudes with characteristic frequencies obtained using envelope (ENV) for vibration acceleration signal; different load torques and supplying frequencies of Figure 7. Examples of harmonic amplitudes with characteristic frequencies obtained using envelope (ENV) for vibration acceleration signal; different load torques and supplying frequencies of PMSM drive.CC-BY · From the PDF
Figure 8 Schematic diagram of a fault detector or fault classifier based on MLP network with different input signals: after FFT or ENV analysis. Figure 8. Schematic diagram of a fault detector or fault classifier based on MLP network with different input signals: after FFT or ENV analysis.
Figure 8 Schematic diagram of a fault detector or fault classifier based on MLP network with different input signals: after FFT or ENV analysis. Figure 8. Schematic diagram of a fault detector or fault classifier based on MLP network with different input signals: after FFT or ENV analysis.CC-BY · From the PDF
Image 9
CC-BY · From the PDF
Figure 10 Schematic diagram of a fault detector or fault classifier based on RBF network with different input signals: after FFT or ENV analysis. Figure 10. Schematic diagram of a fault detector or fault classifier based on RBF network with different input signals: after FFT or ENV analysis.
Figure 10 Schematic diagram of a fault detector or fault classifier based on RBF network with different input signals: after FFT or ENV analysis. Figure 10. Schematic diagram of a fault detector or fault classifier based on RBF network with different input signals: after FFT or ENV analysis.CC-BY · From the PDF
Figure 11 Waveforms of output values from the fault classifier (type II) for different spread values of RBF network based on input vectors from ENV ( a ) or FFT ( b ) analyses. Figure 11. Waveforms of output values from the fault classifier (type II) for different spread values of RBF network based on input vectors from ENV ( a ) or FFT ( b ) analyses. ( a ) ( b ) Figure 11. Waveforms of output values from the fault classifier (type II) for different spread values of RBF network based ( a ) ( b ) Figure 11. Waveforms of output values from the fault classifier (type II) for different spread values of RBF network based
Figure 11 Waveforms of output values from the fault classifier (type II) for different spread values of RBF network based on input vectors from ENV ( a ) or FFT ( b ) analyses. Figure 11. Waveforms of output values from the fault classifier (type II) for different spread values of RBF network based on input vectors from ENV ( a ) or FFT ( b ) analyses. ( a ) ( b ) Figure 11. Waveforms of output values from the fault classifier (type II) for different spread values of RBF network based ( a ) ( b ) Figure 11. Waveforms of output values from the fault classifier (type II) for different spread values of RBF network basedCC-BY · From the PDF
Figure 12 Influence of the number of radial neurons in the hidden layer on the effectiveness of the RBF-based fault detector Figure 12. Influence of the number of radial neurons in the hidden layer on the effectiveness of the RBF-based fault detector for different spread values and input from FFT analysis and σ = 0.1 ( a ) , σ = 0.5 ( b ), ENV analysis and σ = 0.9 ( c ) , σ = 3 ( d ). Figure 12. Influence of the number of radial neurons in the hidden layer on the effectiveness of the RBF-based fault detector for different spread values and input from FFT analysis and s = 0.1 ( a ) , s = 0.5 ( b ), ENV analysis and s = 0.9 ( c ) , s = 3 ( d ). Figure 12. Influence of the number of radial neurons in the hidden layer on the effectiveness of the RBF-based fault detector for different spread values and input from FFT analysis and σ = 0.1 ( a ) , σ = 0.5 ( b ), ENV analysis and σ = 0.9 ( c ) , σ = 3 ( d ).
Figure 12 Influence of the number of radial neurons in the hidden layer on the effectiveness of the RBF-based fault detector Figure 12. Influence of the number of radial neurons in the hidden layer on the effectiveness of the RBF-based fault detector for different spread values and input from FFT analysis and σ = 0.1 ( a ) , σ = 0.5 ( b ), ENV analysis and σ = 0.9 ( c ) , σ = 3 ( d ). Figure 12. Influence of the number of radial neurons in the hidden layer on the effectiveness of the RBF-based fault detector for different spread values and input from FFT analysis and s = 0.1 ( a ) , s = 0.5 ( b ), ENV analysis and s = 0.9 ( c ) , s = 3 ( d ). Figure 12. Influence of the number of radial neurons in the hidden layer on the effectiveness of the RBF-based fault detector for different spread values and input from FFT analysis and σ = 0.1 ( a ) , σ = 0.5 ( b ), ENV analysis and σ = 0.9 ( c ) , σ = 3 ( d ).CC-BY · From the PDF
Figure 13 General schematic diagram of the self-organizing map (SOM).
Figure 13 General schematic diagram of the self-organizing map (SOM).CC-BY · From the PDF
Image 14
CC-BY · From the PDF
Figure 14 Results of the SOM-based fault detector for the 10 × 10 hexagonal map and the linkdist distance function for different values of the neighbourhood radius: 2 ( a ), 6 ( b ), and for different distance function: linkdist ( c ), dist ( d ) with the neighbourhood radius equal to 2. Figure 14. Results of the SOM-based fault detector for the 10 × 10 hexagonal map and the linkdist distance function for different values of the neighbourhood radius: 2 ( a ), 6 ( b ), and for different distance function: linkdist ( c ), dist ( d ) with the neighbourhood radius equal to 2. Figure 14. Results of the SOM-based fault detector for the 10 × 10 hexagonal map and the linkdist distance function for different values of the neighbourhood radius: 2 ( a ), 6 ( b ), and for different distance function: linkdist ( c ), dist ( d ) with the neighbourhood radius equal to 2.
Figure 14 Results of the SOM-based fault detector for the 10 × 10 hexagonal map and the linkdist distance function for different values of the neighbourhood radius: 2 ( a ), 6 ( b ), and for different distance function: linkdist ( c ), dist ( d ) with the neighbourhood radius equal to 2. Figure 14. Results of the SOM-based fault detector for the 10 × 10 hexagonal map and the linkdist distance function for different values of the neighbourhood radius: 2 ( a ), 6 ( b ), and for different distance function: linkdist ( c ), dist ( d ) with the neighbourhood radius equal to 2. Figure 14. Results of the SOM-based fault detector for the 10 × 10 hexagonal map and the linkdist distance function for different values of the neighbourhood radius: 2 ( a ), 6 ( b ), and for different distance function: linkdist ( c ), dist ( d ) with the neighbourhood radius equal to 2.CC-BY · From the PDF
Figure 16 Active areas of the 10 × 10 Kohonen map in the case of an undamaged bearing ( a ) and with a fault ( b ) for the neighbourhood radius equal to 2 and the linkdist function. Figure 16. Active areas of the 10 × 10 Kohonen map in the case of an undamaged bearing ( a ) and with a fault ( b ) for the neighbourhood radius equal to 2 and the linkdist function.
Figure 16 Active areas of the 10 × 10 Kohonen map in the case of an undamaged bearing ( a ) and with a fault ( b ) for the neighbourhood radius equal to 2 and the linkdist function. Figure 16. Active areas of the 10 × 10 Kohonen map in the case of an undamaged bearing ( a ) and with a fault ( b ) for the neighbourhood radius equal to 2 and the linkdist function.CC-BY · From the PDF
Image 17
CC-BY · From the PDF
Figure 17 The results of the SOM classifier for the 10 × 10 hexagonal map and the linkdist distance function for different values of the neighbourhood radius: 2 ( a ), 6 ( b ), and for different distance function: linkdist ( c ), dist ( d ) with the neighbourhood radius equal to 2. Figure 17. The results of the SOM classifier for the 10 × 10 hexagonal map and the linkdist distance function for different values of the neighbourhood radius: 2 ( a ), 6 ( b ), and for different distance function: linkdist ( c ), dist ( d ) with the neighbourhood radius equal to 2. Figure 17. The results of the SOM classifier for the 10 × 10 hexagonal map and the linkdist distance function for different values of the neighbourhood radius: 2 ( a ), 6 ( b ), and for different distance function: linkdist ( c ), dist ( d ) with the neighbourhood radius equal to 2.
Figure 17 The results of the SOM classifier for the 10 × 10 hexagonal map and the linkdist distance function for different values of the neighbourhood radius: 2 ( a ), 6 ( b ), and for different distance function: linkdist ( c ), dist ( d ) with the neighbourhood radius equal to 2. Figure 17. The results of the SOM classifier for the 10 × 10 hexagonal map and the linkdist distance function for different values of the neighbourhood radius: 2 ( a ), 6 ( b ), and for different distance function: linkdist ( c ), dist ( d ) with the neighbourhood radius equal to 2. Figure 17. The results of the SOM classifier for the 10 × 10 hexagonal map and the linkdist distance function for different values of the neighbourhood radius: 2 ( a ), 6 ( b ), and for different distance function: linkdist ( c ), dist ( d ) with the neighbourhood radius equal to 2.CC-BY · From the PDF
Figure 19 Active areas of 10 × 10 SOM map in the case of healthy bearing ( a ), damage to outer raceway ( b ), inner raceway ( c ), rolling element ( d ) for the neighbourhood radius equal to 2 and the linkdist distance function. Figure 19. Active areas of 10 × 10 SOM map in the case of healthy bearing ( a ), damage to outer raceway ( b ), inner raceway ( c ), rolling element ( d ) for the neighbourhood radius equal to 2 and the linkdist distance function.
Figure 19 Active areas of 10 × 10 SOM map in the case of healthy bearing ( a ), damage to outer raceway ( b ), inner raceway ( c ), rolling element ( d ) for the neighbourhood radius equal to 2 and the linkdist distance function. Figure 19. Active areas of 10 × 10 SOM map in the case of healthy bearing ( a ), damage to outer raceway ( b ), inner raceway ( c ), rolling element ( d ) for the neighbourhood radius equal to 2 and the linkdist distance function.CC-BY · From the PDF

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 48 references

Research works printed but not linked: 3

  1. Radu, C. The Most Common Causes of Bearing Failure and the Importance of Bearing Lubrication. RKB Technical Review. Febru- ary 2010, pp. 1-7. Available online: https://www.rkbbearings.com/en/publications.php#sec12 (accessed on 23 December 2020).
  2. Bearing Damage and Failure Analysis, SKF Group, PUB BU/I3 14219/2 EN. June 2017, pp. 1-106. Available online: https: //www.skf.com/binaries/pub12/Images/0901d1968064c148-Bearing-failures---14219_2-EN_tcm_12-297619.pdf (accessed on 23 December 2020).
  3. Perner, P., Ed.; Lecture Notes in Computer Science; Springer: Cham, Switzerland, 2016; Volume 9728. [CrossRef]

Cited by

See all 49

Metadata

Authors
Ewert, P., Orlowska-Kowalska, T., & Jankowska, K.
Affiliations
  • Wrocław University of Science and Technology (PL)
  • Department of Electrical Machines, Drives and Measurements, Wroclaw University of Science and Technology, Wybrzeze Wyspianskiego 27, 50-370 Wroclaw, Poland
Corresponding author
Ewert, Pawel
Venue
Energies
Publisher
MDPI AG
Published
2021-01-30
Volume
14
Issue
3
Pages
712
ISSN
1996-1073 (ISSN-L)
OpenAlex
W3127815279
Cited by
48
References
48
Citations by year
2026: 4 · 2025: 6 · 2024: 10 · 2023: 12 · 2022: 8 · 2021: 8
FWCI
4.23
Type
journal article
Language
en
Open access
Locations
  • Energies (publishedVersion, cc-by)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
Narodowe Centrum Nauki (2017/27/B/ST7/00816)
Topics
  • Physical Sciences › Engineering › Control and Systems Engineering › Machine Fault Diagnosis Techniques
  • Physical Sciences › Engineering › Control and Systems Engineering › Fault Detection and Control Systems
  • Physical Sciences › Engineering › Mechanical Engineering › Gear and Bearing Dynamics Analysis
Keywords
permanent magnet synchronous motor, rolling bearing fault detection, vibration analysis, shallow neural networks, envelope analysis, Hilbert transform, multilayer perceptron, radial basis function neural network, self-organizing maps, electric vehicle drives

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.