A Comparative Study of Semi-Supervised Anomaly Detection Methods for Machine Fault Detection

##plugins.themes.bootstrap3.article.main##

##plugins.themes.bootstrap3.article.sidebar##

Published Jun 27, 2024
Dhiraj Neupane Mohamed Reda Bouadjenek Richard Dazeley Sunil Aryal

Abstract

Industrial automation has extended machines’ runtime, thereby raising breakdown risks. Machine breakdowns not only have economic and productivity consequences, but they can also be fatal. Thus, the early detection of fault signs is essential for the safe and uninterrupted operation of machinery and its maintenance. In the last few years, machine learning has been widely used in machine condition monitoring. Most existing approaches rely on supervised learning techniques, which face challenges in real-world scenarios due to the lack of enough labelled fault data. Additionally, models trained on historical fault data might struggle to detect new and unseen faults accurately in the future. Therefore, this research uses semi-supervised Anomaly Detection (AD) techniques to detect abnormal patterns in machines’ vibration signals. As semi-supervised techniques are trained on normal data only, they do not require faulty samples and abnormal patterns are detected based on their deviations from the learned normal pattern. We compared the effectiveness of seven state-of-the-art AD methods, ranging from traditional approaches such as isolation forest and local outlier factor to more recent Deep Learning (DL) approaches based on autoencoders. We evaluated the effectiveness of different feature types extracted from the raw vibration signals, including simple statistical features like kurtosis, mean, peak-to-peak, and more complex representations like the scalogram images. Our study on three public datasets, with unique challenges, shows that the traditional methods based on simple statistical analysis have shown comparable and sometimes superior performance to more complex DL approaches. The use of traditional approaches offers simplicity and lower computational needs. Thus, our study recommends that future researchers start with the traditional approaches first and then jump to DL methods if necessary.

How to Cite

Neupane, D., Bouadjenek, M. R., Dazeley, R. ., & Aryal, S. (2024). A Comparative Study of Semi-Supervised Anomaly Detection Methods for Machine Fault Detection. PHM Society European Conference, 8(1), 10. https://doi.org/10.36001/phme.2024.v8i1.4053
Abstract 418 | PDF Downloads 329

##plugins.themes.bootstrap3.article.details##

Keywords

Semi-Supervised Learning, Anomaly Detection, Machinery Fault, Machine Learning, Deep Learning, Machine Health Monitoring

References
@inproceedings{ahmad2020autoencoder,
title={Autoencoder-based condition monitoring and anomaly detection method for rotating machines},
author={Ahmad, Sabtain and Styp-Rekowski, Kevin and Nedelkoski, Sasho and Kao, Odej},
booktitle={2020 IEEE International Conference on Big Data (Big Data)},
pages={4093--4102},
year={2020},
organization={IEEE}
}

@article{aryal2021usfad,
title={usfAD: a robust anomaly detector based on unsupervised stochastic forest},
author={Aryal, Sunil and Santosh, KC and Dazeley, Richard},
journal={International Journal of Machine Learning and Cybernetics},
volume={12},
pages={1137--1150},
year={2021},
publisher={Springer}
}

@article{audibert2022deep,
title={Do deep neural networks contribute to multivariate time series anomaly detection?},
author={Audibert, Julien and Michiardi, Pietro and Guyard, Fr{\'e}d{\'e}ric and Marti, S{\'e}bastien and Zuluaga, Maria A},
journal={Pattern Recognition},
volume={132},
pages={108945},
year={2022},
publisher={Elsevier}
}

@inproceedings{breunig2000lof,
title={LOF: identifying density-based local outliers},
author={Breunig, Markus M and Kriegel, Hans-Peter and Ng, Raymond T and Sander, J{\"o}rg},
booktitle={Proceedings of the 2000 ACM SIGMOD international conference on Management of data},
pages={93--104},
year={2000}
}

@article{chaleshtori2024novel,
title={A novel bearing fault diagnosis approach using the Gaussian mixture model and the weighted principal component analysis},
author={Chaleshtori, Amir Eshaghi and Aghaie, Abdollah},
journal={Reliability Engineering \& System Safety},
volume={242},
pages={109720},
year={2024},
publisher={Elsevier}
}

@article{das2023machineFaultComprehensiveREview2023,
title={Machine learning for fault analysis in rotating machinery: A comprehensive review},
author={Das, Oguzhan and Das, Duygu Bagci and Birant, Derya},
journal={Heliyon},
year={2023},
publisher={Elsevier}
}

@article{guo2020bearing,
title={Bearing intelligent fault diagnosis based on wavelet transform and convolutional neural network},
author={Guo, Junfeng and Liu, Xingyu and Li, Shuangxue and Wang, Zhiming},
journal={Shock and Vibration},
volume={2020},
pages={1--14},
year={2020},
publisher={Hindawi Limited}
}

@inproceedings{he2016deepResNet,
title={Deep residual learning for image recognition},
author={He, Kaiming and Zhang, Xiangyu and Ren, Shaoqing and Sun, Jian},
booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition},
pages={770--778},
year={2016}
}

@article{jiao2019hierarchicalrotaingMachinery,
title={Hierarchical discriminating sparse coding for weak fault feature extraction of rolling bearings},
author={Jiao, Jinyang and Zhao, Ming and Lin, Jing and Liang, Kaixuan},
journal={Reliability Engineering \& System Safety},
volume={184},
pages={41--54},
year={2019},
publisher={Elsevier}
}

@inproceedings{kumagai2021semi,
title={Semi-supervised anomaly detection on attributed graphs},
author={Kumagai, Atsutoshi and Iwata, Tomoharu and Fujiwara, Yasuhiro},
booktitle={2021 International Joint Conference on Neural Networks (IJCNN)},
pages={1--8},
year={2021},
organization={IEEE}
}

@inproceedings{lessmeier2016condition,
title={Condition monitoring of bearing damage in electromechanical drive systems by using motor current signals of electric motors: A benchmark data set for data-driven classification},
author={Lessmeier, Christian and Kimotho, James Kuria and Zimmer, Detmar and Sextro, Walter},
booktitle={PHM Society European Conference},
volume={3},
number={1},
year={2016}
}

@article{li2020systematicTransferReview,
title={A systematic review of deep transfer learning for machinery fault diagnosis},
author={Li, Chuan and Zhang, Shaohui and Qin, Yi and Estupinan, Edgar},
journal={Neurocomputing},
volume={407},
pages={121--135},
year={2020},
publisher={Elsevier}
}

@INPROCEEDINGS{4781136isolationForest,
author={Liu, Fei Tony and Ting, Kai Ming and Zhou, Zhi-Hua},
booktitle={2008 Eighth IEEE International Conference on Data Mining},
title={Isolation Forest},
year={2008},
volume={},
number={},
pages={413-422},
doi={10.1109/ICDM.2008.17}
}

@article{neupane2024data,
title={Data-driven Machinery Fault Detection: A Comprehensive Review},
author={Neupane, Dhiraj and Bouadjenek, Mohamed Reda and Dazeley, Richard and Aryal, Sunil},
journal={arXiv preprint arXiv:2405.18843},
year={2024}
}

@ARTICLE{dhirajSN_CNN,
author={Neupane, Dhiraj and Kim, Yunsu and Seok, Jongwon},
journal={IEEE Access},
title={Bearing Fault Detection Using Scalogram and Switchable Normalization-Based CNN (SN-CNN)},
year={2021},
volume={9},
number={},
pages={88151-88166},
doi={10.1109/ACCESS.2021.3089698}
}

@article{dhirajCwru,
author = {Dhiraj Neupane and Jongwon Seok},
title = {Bearing Fault Detection and Diagnosis Using Case Western Reserve University Dataset With Deep Learning Approaches: A Review},
journal = {IEEE Access},
volume = {8},
pages = {93155--93178},
year = {2020},
doi = {10.1109/ACCESS.2020.2990528}
}

@article{peeters2024fatigue,
title={Fatigue crack detection in planetary gears: Insights from the HUMS2023 data challenge},
author={Peeters, C{\'e}dric and Wang, Wenyi and Blunt, David and Verstraeten, Timothy and Helsen, Jan},
journal={Mechanical Systems and Signal Processing},
volume={212},
pages={111292},
year={2024},
publisher={Elsevier}
}

@article{pourhabibi2020fraud,
title={Fraud detection: A systematic literature review of graph-based anomaly detection approaches},
author={Pourhabibi, Tahereh and Ong, Kok-Leong and Kam, Booi H and Boo, Yee Ling},
journal={Decision Support Systems},
volume={133},
pages={113303},
year={2020},
publisher={Elsevier}
}

@article{ sawalhi2024helicopter,
title={Helicopter Planet Gear Rim Crack Diagnosis and Trending Using Cepstrum Editing Enhanced with Deconvolution},
author={Sawalhi, Nader and Wang, Wenyi and Blunt, David},
journal={Sensors},
volume={24},
number={8},
pages={2593},
year={2024},
publisher={MDPI}
}

@article{shin2005oneclassSVM,
title={Support vector method for novelty detection},
author={Sch{\"o}lkopf, Bernhard and Williamson, Robert C and Smola, Alex and Shawe-Taylor, John and Platt, John},
journal={Advances in neural information processing systems},
volume={12},
year={1999}
}

@article{simonyan2014veryVGGNet,
title={Very deep convolutional networks for large-scale image recognition},
author={Simonyan, Karen and Zisserman, Andrew},
journal={arXiv preprint arXiv:1409.1556},
year={2014}
}

@article{torabi2023practical,
title={Practical autoencoder based anomaly detection by using vector reconstruction error},
author={Torabi, Hasan and Mirtaheri, Seyedeh Leili and Greco, Sergio},
journal={Cybersecurity},
volume={6},
number={1},
pages={1},
year={2023},
publisher={Springer}
}

@article{vos2022vibration,
title={Vibration-based anomaly detection using LSTM/SVM approaches},
author={Vos, Kilian and Peng, Zhongxiao and Jenkins, Christopher and Shahriar, Md Rifat and Borghesani, Pietro and Wang, Wenyi},
journal={Mechanical Systems and Signal Processing},
volume={169},
pages={108752},
year={2022},
publisher={Elsevier}
}

@misc{wang2023helicopter,
title={Helicopter Main Gearbox Planet Gear Crack Propagation Test Dataset},
author={Wang, Wenyi and Blunt, David and Kappas, J},
year={2023}
}

@article{wang2023deep,
title={Is deep learning superior to traditional techniques in machine health monitoring applications},
author={Wang, W and Vos, K and Taylor, J and Jenkins, C and Bala, B and Whitehead, L and Peng, Z},
journal={The Aeronautical Journal},
volume={127},
number={1318},
pages={2105--2117},
year={2023},
publisher={Cambridge University Press}
}

@article{xie2024robust,
title={A Robust Anomaly Detection Model for Pumps Based on the Spectral Residual With Self-Attention Variational Autoencoder},
author={Xie, Tianming and Xu, Qifa and Jiang, Cuixia and Gao, Zhiwei and Wang, Xiangxiang},
journal={IEEE Transactions on Industrial Informatics},
year={2024},
publisher={IEEE}
}

@ARTICLE{9270010,
author={Zhang, Shen and Ye, Fei and Wang, Bingnan and Habetler, Thomas G.},
journal={IEEE Sensors Journal},
title={Semi-Supervised Bearing Fault Diagnosis and Classification Using Variational Autoencoder-Based Deep Generative Models},
year={2021},
volume={21},
number={5},
pages={6476-6486},
doi={10.1109/JSEN.2020.3040696}
}

@INPROCEEDINGS{9291099,
author={Zhang, Shen and Ye, Fei and Wang, Bingnan and Habetler, Thomas G.},
booktitle={2020 23rd International Conference on Electrical Machines and Systems (ICEMS)},
title={Few-Shot Bearing Anomaly Detection via Model-Agnostic Meta-Learning},
year={2020},
volume={},
number={},
pages={1341-1346},
doi={10.23919/ICEMS50442.2020.9291099}
}

@article{zhang2019semi,
title={Semi-supervised learning of bearing anomaly detection via deep variational autoencoders},
author={Zhang, Shen and Ye, Fei and Wang, Bingnan and Habetler, Thomas G},
journal={arXiv preprint arXiv:1912.01096},
year={2019}
}



@article{zheng2002gear,
title={Gear fault diagnosis based on continuous wavelet transform},
author={Zheng, Haibao and Li, Zhiyuan and Chen, Xinzhao},
journal={Mechanical systems and signal processing},
volume={16},
number={2-3},
pages={447--457},
year={2002},
publisher={Elsevier}
}


@Article{nuclear2023,
author = "Xianping Zhong and Lin Zhang and Heng Ban",
title = "Deep reinforcement learning for class imbalance fault diagnosis of equipment in nuclear power plants",
journal = "Annals of Nuclear Energy",
volume = "184",
number = "",
month = may,
year = "2023",
pages = "109685",
doi = "10.1016/j.anucene.2023.109685"
}

@Article{machines10070515,
AUTHOR = {Zong, Xia and Yang, Rui and Wang, Hongshu and Du, Minghao and You, Pengfei and Wang, Su and Su, Hao},
TITLE = {Semi-Supervised Transfer Learning Method for Bearing Fault Diagnosis with Imbalanced Data},
JOURNAL = {Machines},
VOLUME = {10},
YEAR = {2022},
NUMBER = {7},
ARTICLE-NUMBER = {515},
ISSN = {2075-1702},
DOI = {10.3390/machines10070515}
}
Section
Technical Papers