Principal fault localization of the faults chain, as a branch of fault diagnosis in wind turbine system, has been an essential problem to ensure the reliability and security in the real wind farms recently. It can be solved by machine learning techniques with historical data labeled with principal faults. However, most real data are unlabeled, since the labeled is expensive to obtain, which increases the difficulty to localize the principal fault if just using unlabeled data and few labeled data. So, in this article, a novel approach using unlabeled data is proposed for principal fault localization of the faults chain in wind turbine systems. First, a deep learning model, stacked sparse autoencoders, is introduced to learn and extract high-level features from data. Then, we present a graphical temporal semi-supervised learning algorithm to develop the pseudo-labeled data set with an unlabeled data set. Considering the temporal correlation of wind power data, we add a time weight vector and apply the cosine-similarity in the proposed algorithm. Finally, based on the pseudo-labeled data set, a classifier model is built and trained for the principal fault localization of the faults chain. The proposed approach is verified by the real buffer data set collected from two wind farms in China, and the experimental results show its effectiveness in practice.
ChehouriAYounesRIlincaA, et al. Review of performance optimization techniques applied to wind turbines. Appl Energy2015; 142: 361–388.
2.
KabirMOoAMRabbaniM. A brief review on offshore wind turbine fault detection and recent development in condition monitoring based maintenance system. In: 2015 Australasian universities power engineering conference (AUPEC), Wollongong, NSW, 27–30 September, pp.1–7. New York: IEEE.
3.
BrišRByczanskiPGoňoR, et al. Discrete maintenance optimization of complex multi-component systems. Reliabil Eng Syst Safety2017; 168: 80–89.
4.
LiuRYangBZioE, et al. Artificial intelligence for fault diagnosis of rotating machinery: a review. Mech Syst Sig Process2018; 108: 33–47.
5.
LiuBWuJYaoL, et al. Ontology-based fault diagnosis: a decade in review. In: Proceedings of the 11th international conference on computer modeling and simulation, Melbourne, VIC, Australia, 16–18 January 2019, pp.112–116. New York: ACM Press.
6.
GaoZDingSXCecatiC.Real-time fault diagnosis and fault-tolerant control. IEEE Trans Indus Electron2015; 62(6): 3752–3756.
7.
BeddekKMerabetAKesraouiM, et al. Signal-based sensor fault detection and isolation for PMSG in wind energy conversion systems. IEEE Trans Instrument Measure2017; 66(9): 2403–2412.
8.
YangRLiHHeC, et al. Rolling element bearing weak fault diagnosis based on optimal wavelet scale cyclic frequency extraction. Proc IMechE, Part I: J Systems and Control Engineering2018; 232(7): 895–908.
9.
JafariHPoshtanJSadeghiH.Application of fuzzy data fusion theory in fault diagnosis of rotating machinery. Proc IMechE, Part I: J Systems and Control Engineering2018; 232(8): 1015–1024.
10.
BadihiHZhangYHongH.Fault-tolerant cooperative control in an offshore wind farm using model-free and model-based fault detection and diagnosis approaches. Appl Energy2017; 201: 284–307.
11.
ChoSGaoZMoanT.Model-based fault detection, fault isolation and fault-tolerant control of a blade pitch system in floating wind turbines. Renewab Energy2018; 120: 306–321.
12.
AsgariSYazdizadehA.Robust model-based fault diagnosis of mechanical drive train in v47/660 kw wind turbine. Energy Syst2018; 9(4): 921–952.
13.
ZhangJJiangNLiH, et al. Online health assessment of wind turbine based on operational condition recognition. Trans Instit Measure Contr2019; 41(10): 2970–2981.
14.
SunPLiJWangC, et al. A generalized model for wind turbine anomaly identification based on scada data. Appl Energy2016; 168: 550–567.
15.
LeahyKHuRLKonstantakopoulosIC, et al. Diagnosing and predicting wind turbine faults from scada data using support vector machines. Int J Prognost Health Manage2018; 9(1): 1–11.
16.
QianPMaXCrossP.Integrated data-driven model-based approach to condition monitoring of the wind turbine gearbox. IET Renewab Power Generat2017; 11(9): 1177–1185.
17.
LeiYJiaFLinJ, et al. An intelligent fault diagnosis method using unsupervised feature learning towards mechanical big data. IEEE Trans Indus Electron2016; 63(5): 3137–3147.
18.
WangLZhangZXuJ, et al. Wind turbine blade breakage monitoring with deep autoencoders. IEEE Trans Smart Grid2018; 9(4): 2824–2833.
19.
JiangGHeHXieP, et al. Stacked multilevel-denoising autoencoders: a new representation learning approach for wind turbine gearbox fault diagnosis. IEEE Trans Instrument Measure2017; 66(9): 2391–2402.
20.
ShaoHJiangHZhangH, et al. Electric locomotive bearing fault diagnosis using novel convolutional deep belief network. IEEE Trans Indus Electron2018; 65(3): 2727–2736.
21.
YaoZYuYYaoJ.Artificial neural network–based internal leakage fault detection for hydraulic actuators: an experimental investigation. Proc IMechE, Part I: J Systems and Control Engineering2018; 232(4): 369–382.
ShanYZhouJJiangW, et al. A fault diagnosis method for rotating machinery based on improved variational mode decomposition and a hybrid artificial sheep algorithm. Measure Sci Technol2019; 30(5): 055002.
24.
QiaoWLuD.A survey on wind turbine condition monitoring and fault diagnosis—part I: components and subsystems. IEEE Trans Indus Electron2015; 62(10): 6536–6545.
25.
ZhuX.Semi-supervised learning literature survey. Madison, WI: Computer Science, University of Wisconsin-Madison, 2006.
26.
AbdelgayedTSMorsiWGSidhuTS.Fault detection and classification based on co-training of semisupervised machine learning. IEEE Trans Indus Electron2018; 65(2): 1595–1605.
27.
Razavi-FarRHallajiEFarajzadeh-ZanjaniM, et al. Information fusion and semi-supervised deep learning scheme for diagnosing gear faults in induction machine systems. IEEE Trans Indus Electron2018; 66(8): 6331–6342.
28.
DuBTangXWangZ, et al. Robust graph-based semisupervised learning for noisy labeled data via maximum correntropy criterion. IEEE Trans Cybernet2018; 49(4): 1440–1453.
29.
TuYLinYWangJ, et al. Semi-supervised learning with generative adversarial networks on digital signal modulation classification. Comput Mater Continua2018; 55(2): 243–254.
30.
LiQHanZWuXM. Deeper insights into graph convolutional networks for semi-supervised learning. In: Proceedings of the 32nd AAAI conference on artificial intelligence, New Orleans, LA, 2–7 February 2018. Menlo Park, CA: Association for the Advancement of Artificial Intelligence.
31.
AppiceAGuccionePMalerbaD.A novel spectral-spatial co-training algorithm for the transductive classification of hyperspectral imagery data. Pattern Recognition2017; 63: 229–245.
32.
ZhouDBousquetOLalTN, et al. Learning with local and global consistency. In: ThrunTSaulLKSchölkopfB (eds) Advances in neural information processing systems, vol. 16. Cambridge, MA: MIT Press, pp.321–328.
33.
BerthelotDCarliniNGoodfellowI, et al. Mixmatch: a holistic approach to semi-supervised learning, 2019, https://arxiv.org/abs/1905.02249
34.
HanJPeiJKamberM.Data mining: concepts and techniques. Amsterdam: Elsevier, 2011.
BengioYLamblinPPopoviciD, et al. Greedy layer-wise training of deep networks. In: SchölkopfBPlattJCHoffmanT (eds) Advances in neural information processing systems, vol. 19. Cambridge, MA: MIT Press, pp.153–160.
AminiMRUsunierN.Learning with partially labeled and interdependent data. Berlin: Springer, 2015.
40.
BhattacharyyaA.On a measure of divergence between two multinomial populations. Sankhyā1946; 7(4): 401–406.
41.
SaltonGMcGillMJ.Introduction to modern information retrieval. New York: McGraw Hill, 1986.
42.
HintonGEOsinderoSTehYW.A fast learning algorithm for deep belief nets. Neural Comput2006; 18(7): 1527–1554.
43.
GlorotXBengioY. Understanding the difficulty of training deep feedforward neural networks. In: Proceedings of the 13th international conference on artificial intelligence and statistics, vol. 9, Chia Laguna Resort, Sardinia, Italy, 13–15 May 2010, pp.249–256. JMLR W&CP 9.
