Hostname: page-component-76fb5796d-5g6vh Total loading time: 0 Render date: 2024-04-25T20:54:41.106Z Has data issue: false hasContentIssue false
  • English
  • Français

Is deep learning superior to traditional techniques in machine health monitoring applications

Part of: The 19th and 20th Australian International Aerospace Congresses The AJ Special Rotorcraft Collection

Published online by Cambridge University Press:  14 August 2023

W. Wang Open the ORCID record for W. Wang [Opens in a new window] ,
K. Vos ,
J. Taylor ,
C. Jenkins ,
B. Bala ,
L. Whitehead  and
Z. Peng
W. Wang*
Affiliation:
Defence Science and Technology Group, Aerospace Division, Fishermans Bend, VIC, Australia
K. Vos
Affiliation:
University of New South Wales, Sydney, NSW, Australia
J. Taylor
Affiliation:
Defence Science and Technology Group, Research Services Division, Fairbairn, ACT, Australia
C. Jenkins
Affiliation:
University of New South Wales, Sydney, NSW, Australia
B. Bala
Affiliation:
Defence Science and Technology Group, Research Services Division, Fairbairn, ACT, Australia
L. Whitehead
Affiliation:
Defence Science and Technology Group, Aerospace Division, Fishermans Bend, VIC, Australia
Z. Peng
Affiliation:
University of New South Wales, Sydney, NSW, Australia
*
Corresponding author: W. Wang; Email: wenyi.wang@defence.gov.au
Get access Rights & Permissions [Opens in a new window]

Abstract

In recent years, there has been significant momentum in applying deep learning (DL) to machine health monitoring (MHM). It has been widely claimed that DL methodologies are superior to more traditional techniques in this area. This paper aims to investigate this claim by analysing a real-world dataset of helicopter sensor faults provided by Airbus. Specifically, we will address the problem of machine sensor health unsupervised classification. In a 2019 worldwide competition hosted by Airbus, Fujitsu Systems Europe (FSE) won first prize by achieving an F1-score of 93% using a DL model based on generative adversarial networks (GAN). In another comprehensive study, various modified and existing image encoding methods were compared for the convolutional auto-encoder (CAE) model. The best classification result was achieved using the scalogram as the image encoding method, with an F1-score of 91%. In this paper, we use these two studies as benchmarks to compare with basic statistical analysis methods and the one-class supporting vector machine (SVM). Our comparative study demonstrates that while DL-based techniques have great potential, they are not always superior to traditional methods. We therefore recommend that all future published studies of applying DL methods to MHM include appropriately selected traditional reference methods, wherever possible.

Keywords

faulty-sensor detectiontime sequence classificationmachine learningdeep learningstatistical signal analysismachine health monitoring
Type
Research Article
Information
The Aeronautical Journal , Volume 127 , Special Issue 1318: AIAC 2023 , December 2023 , pp. 2105 - 2117
Check for updates
Copyright
© The Author(s), 2023. Published by Cambridge University Press on behalf of Royal Aeronautical Society

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

Choi, K., Yi, J., Park, C. and Yoon, S. Deep learning for anomaly detection in time-series data: review, analysis, and guidelines, IEEE Access, 2021, 9. https://doi.org/10.1109/ACCESS.2021.3107975 Google Scholar
Zhao, R., Yan, R., Chen, Z., Mao, K., Wang, P. and Gao, R.X. Deep learning and its applications to machine health monitoring, Mech Syst Signal Process., 2019, 115, pp 213–237. https://doi.org/10.1016/j.ymssp.2018.05.050 CrossRefGoogle Scholar
Wang, W., Taylor, J. and Rees, R. Recent advancement of deep learning applications to machine condition monitoring part 1: a critical review, Acoust. Aust., 2021, 49, pp 207–219. https://doi.org/10.1007/s40857-021-00222-9 CrossRefGoogle Scholar
Li, D., Chen, D., Jin, B., Shi, L., Goh, J. and Ng, S.K. MAD-GAN: multivariate anomaly detection for time series data with generative adversarial networks, Proceedings of International Conference on Artificial Neural Networks – ICANN 2019: Artificial Neural Networks and Machine Learning: Text and Time Series. Lecture Notes in Computer Science, vol. 11730, Springer, Cham. https://doi.org/10.1007/978-3-030-30490-4_56 CrossRefGoogle Scholar
Garcia, G.R., Michau, G., Ducoffe, M., Gupta, J.S. and Fink, O. Time series to images: monitoring the condition of industrial assets with deep learning image processing algorithms, 2020. arXiv preprint. https://arxiv.org/abs/2005.07031v2 Google Scholar
Garcia, G.R., Michau, G., Ducoffe, M., Gupta, J.S. and Fink, O. Temporal signals to images: monitoring the condition of industrial assets with deep learning image processing algorithms, Proc. Inst. Mech. Eng. Part O J. Risk Reliab., 2022, 236, (4), pp 617–627. https://doi.org/10.1177/1748006x21994446 Google Scholar
Zhao, R., Yan, R., Wang, J. and Mao, K. Learning to monitor machine health with convolutional bi-Directional LSTM networks, Sensors, 2017, 17, (273). https://doi.org/10.3390/s17020273 Google ScholarPubMed
Li, X., Zhang, W. and Ding, Q. Deep learning-based remaining useful life estimation of bearings using multi-scale feature extraction, Reliab. Eng. Syst. Safety, 2019, 182, pp 208–218. https://doi.org/10.1016/j.ress.2018.11.011 CrossRefGoogle Scholar
Shao, S., McAleer, S., Yan, R. and Baldi, P. Highly accurate machine fault diagnosis using deep transfer learning, IEEE Trans. Ind. Inf., 2019, 15, (4), pp 2446–2455. https://doi.org/10.1109/TII.2018.2864759 CrossRefGoogle Scholar
Cortes, C. and Vapnik, V. Support-vector networks, Mach. Learn., 1995, 20, pp 273–297.CrossRefGoogle Scholar
Fernández-Francos, D., Marténez-Rego, D., Fontenla-Romero, O. and Alonso-Betanzos, A. Automatic bearing fault diagnosis based on one-class m-SVM, Comput. Ind. Eng., 2013, 64, pp 357–365. https://doi.org/10.1016/j.cie.2012.10.013 CrossRefGoogle Scholar
Widodo, A. and Yang, B.S. Support vector machine in machine condition monitoring and fault diagnosis, Mech. Syst. Signal Process., 2007, 21, pp 2560–2574. https://doi.org/10.1016/j.ymssp.2006.12.007 CrossRefGoogle Scholar
Bi, J., Bennett, K., Embrechts, M., Breneman, C. and Song, M. Dimensionality reduction via sparse support vector machines, J. Mach. Learn. Res., 2003, 3, pp 1229–1243.Google Scholar
Wang, W., Taylor, J. and Rees, R. Recent advancement of deep learning applications to machine condition monitoring part 2: supplement views and a case study, Acoust. Aust., 2021, 49, pp 221–228. https://doi.org/10.1007/s40857-021-00235-4 CrossRefGoogle Scholar