Deep Learning Method for Fault Detection of Wind Turbine Converter
Source abstract
Abstract
The converter is an important component in wind turbine power drive-train systems, and usually, it has a higher failure rate. Therefore, detecting the potential faults for prediction of its failure has become indispensable for condition-based maintenance and operation of wind turbines. This paper presents an approach to wind turbine converter fault detection using convolutional neural network models which are developed by using wind turbine Supervisory Control and Data Acquisition (SCADA) system data. The approach starts with the selection of fault indicator variables, and then the fault indicator variables data are extracted from a wind turbine SCADA system. Using the data, radar charts are generated, and the convolutional neural network models are applied to feature extraction from the radar charts and characteristic analysis of the feature for fault detection. Based on the analysis of the Octave Convolution (OctConv) network structure, an improved AOctConv (Attention Octave Convolution) structure is proposed in this paper, and it is applied to the ResNet50 backbone network (named as AOC–ResNet50). It is found that the algorithm based on AOC–ResNet50 overcomes the issues of information asymmetry caused by the asymmetry of the sampling method and the damage to the original features in the high and low frequency domains by the OctConv structure. Finally, the AOC–ResNet50 network is employed for fault detection of the wind turbine converter using 10 min SCADA system data. It is verified that the fault detection accuracy using the AOC–ResNet50 network is up to 98.0%, which is higher than the fault detection accuracy using the ResNet50 and Oct–ResNet50 networks. Therefore, the effectiveness of the AOC–ResNet50 network model in wind turbine converter fault detection is identified. The novelty of this paper lies in a novel AOC–ResNet50 network proposed and its effectiveness in wind turbine fault detection. This was verified through a comparative study on wind turbine power converter fault detection with other competitive convolutional neural network models for deep learning.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Machine Fault Diagnosis Techniques · Fault Detection and Control Systems · Power System Reliability and Maintenance
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
School of Artificial Intelligence, Hebei University of Technology, Tianjin 300131, China
School of Electronic and Control Engineering, North China Institute of Aerospace Engineering, Langfang 065000, China
Faculty of Engineering and Information Sciences, University of Wollongong, Wollongong, NSW 2522, Australia
Department of Technical Development, AT&M Environmental Engineering Technology Co., Ltd., Beijing 100801, China
Funding
National Natural Science Foundation of China · Hebei Natural Science Foundation · Youth Fund of Hebei Education Department · Key Project of North China Institute of Aerospace Engineering
References (61)
Show source-provided references
1. (2020, November 20). WWEA. Available online: https://wwindea.org/blog/category/statistics/.
2. Optimal control of grid-connected converter under asymmetrical grid voltage fault
3. Fault estimation of wind turbines using adaptive and parameter estimation schemes10.1002/acs.2792 ↗
4. A fault detection method for pitch actuators faults in wind turbines
5. A new fault diagnosis approach for the pitch system of wind turbines10.1177/1687814017703350 ↗
6. A survey on wind turbine condition monitoring and fault diagnosis−Part II: Signals and signal processing methods10.1109/TIE.2015.2422394 ↗
7. Automatic fault prediction of wind turbine main bearing based on SCADA data and artificial neural network
8. Data-driven techniques for the fault diagnosis of a wind turbine benchmark10.2478/amcs-2018-0018 ↗
9. Detection and classification of faults in pitch-regulated wind turbine generators using normal behaviour models based on performance curves10.1016/j.renene.2016.12.075 ↗
10. Data driven sensor and actuator fault detection and isolation in wind turbine using classifier fusion10.1016/j.renene.2017.03.051 ↗
11. Anomaly detection and fault analysis of wind turbine components based on deep learning network10.1016/j.renene.2018.05.024 ↗
12. Deep Learning for fault detection in wind turbines10.1016/j.rser.2018.09.012 ↗
13. DNN-based approach for fault detection in a direct drive wind turbine10.1049/iet-rpg.2017.0867 ↗
14. Using SCADA data for wind turbine condition monitoring—A review10.1049/iet-rpg.2016.0248 ↗
15. Fault diagnosis of converter used in wind power generation based on wavelet packet analysis and SVM
16. An SVM-based solution for fault detection in wind turbines10.3390/s150305627 ↗
17. Zhang, P.J., and Lu, D.L. (2019). A survey of condition monitoring and fault diagnosis toward integrated O&M for wind turbines. Energies, 12.10.3390/en12142801 ↗
18. Xiao, C., Liu, Z.J., Zhang, T.L., and Zhang, L. (2019). On fault prediction for wind turbine pitch system using radar chart and support vector machine approach. Energies, 12.10.3390/en12142693 ↗
19. Kavaz, A.G., and Barutcu, B. (2018). Fault detection of wind turbine sensors using artificial neural networks. J. Sens., 5628429.10.1155/2018/5628429 ↗
20. Wind turbine fault detection based on SCADA data analysis using ANN10.1007/s40436-014-0061-6 ↗
21. Kim, K., Parthasarathy, G., Uluyol, O., Foslien, W., Shuangwen, S., and Fleming, P. (2011, January 7–10). Use of SCADA data for failure detection in wind turbines. Proceedings of the 2011 Energy Sustainability Conference and Fuel Cell Conference, Washington, DC, USA.10.1115/ES2011-54243 ↗
22. Fault diagnostic methods for wind turbine: A review
23. Maldonado-Correa, J., Martín-Martínnez, S., Artigao, E., and Gómez-Lázaro, E. (2020). Using SCADA data for wind turbine condition monitoring: A systematic literature review. Energies, 13.10.3390/en13123132 ↗
24. Learning deep representation of imbalanced SCADA data for fault detection of wind turbines10.1016/j.measurement.2019.03.029 ↗
25. Li, M., Wang, S.X., Fang, S.X., and Zhao, J. (2020). Anomaly detection of wind turbines based on deep small-world neural network. Appl. Sci., 10.10.3390/app10041243 ↗
26. Machine learning methods for wind turbine condition monitoring: A review10.1016/j.renene.2018.10.047 ↗
27. Multiscale convolutional neural networks for fault diagnosis of wind turbine gearbox10.1109/TIE.2018.2844805 ↗
28. Wang, Z.J., Zhen, L.K., and Du, W.H. (2019). A novel method for intelligent fault diagnosis of bearing based on Capsule neural network. Complexity, 6943234.10.1155/2019/6943234 ↗
29. Deep learning with knowledge transfer for explainable anomaly prediction in wind turbines10.1002/we.2510 ↗
30. Discriminative dictionary learning based sparse representation classification for intelligent fault identification of planet bearings in wind turbine10.1016/j.renene.2020.01.093 ↗
31. A threshold self-setting condition monitoring scheme for wind turbine generator bearings based on deep convolutional generative adversarial networks10.1016/j.measurement.2020.108234 ↗
32. Gearbox fault diagnosis using a deep learning model with limited data sample10.1109/TII.2020.2967822 ↗
33. Current residual vector-based open-switch fault diagnosis of inverters in PMSM drive systems10.1109/TPEL.2014.2360834 ↗
34. Condition monitoring of wind power converters using heat flux sensor
35. Modeling and control of permanent-magnet synchronous generators under open-switch converter faults10.1109/TPEL.2018.2855423 ↗
36. Fault diagnostic algorithm for three-level neutral point clamped AC motor drives, based on the average current Park’s vector10.1049/iet-pel.2013.0416 ↗
37. Open-circuit faults diagnosis in back-to-back converters of DF wind turbine10.1049/iet-rpg.2016.0150 ↗
38. Converters open-circuit fault-diagnosis methods research for direct-driven permanent magnet wind power system
39. Qiu, Y., Jiang, H., Feng, Y., Cao, M., Zhao, Y., and Li, D. (2016). A new fault diagnosis algorithm for PMSG wind turbine power converters under variable wind speed conditions. Energies, 9.10.3390/en9070548 ↗
40. Design and evaluation of voltage measurement-based sectoral diagnosis method for inverter open switch faults of permanent magnet synchronous motor drives10.1049/iet-epa.2011.0315 ↗
41. A voltage-based approach without extra hardware for open-circuit fault diagnosis in closed-loop PWM AC regenerative drives10.1109/TIE.2013.2279383 ↗
42. Characteristic analysis for permanent magnet synchronous generator wind power systems during converter faults
43. Fault diagnosis of open-circuit faults in converters of direct-driven permanent magnet wind power generation systems based on line voltage errors
44. Switch fault diagnosis method for series-parallel forward DC-DC converter system10.1109/TIE.2018.2866109 ↗
45. Open-circuit fault diagnosis for matrix converter drives and remedial operation using carrier-based modulation methods10.1109/TIE.2013.2240639 ↗
46. Diagnosis of multiple open-circuit switch faults based on Long Short-Term Memory Network for DFIG-based wind turbine systems10.1109/JESTPE.2019.2908981 ↗
47. Hidden Markov model based principal component analysis for intelligent fault diagnosis of wind energy converter systems10.1016/j.renene.2020.01.010 ↗
48. Fault diagnosis of wind power converters based on compressed sensing theory and weight constrained AdaBoost-SVM
49. Liu, Z.J., Xiao, C., Zhang, T.L., and Zhang, X. (2020). Research fault detection for three types of wind turbine subsystems using machine learning. Energies, 13.10.3390/en13020460 ↗
50. SyncedReview (2020, November 30). Upgrading CNN with OctConv. Available online: https://medium.com/syncedreview/upgrading-cnn-with-octconv-5ed9770759be.
51. Tang, Z.Y. (2014). Comprehensive Evaluation of Green Building Design Schemes, Chang’an University.
52. Radar plots: A useful way for presenting multivariate health care data10.1016/j.jclinepi.2007.04.021 ↗
53. Porter, M.M., and Niksiar, P. (2018). Multidimensional mechanics: Performance mapping of natural biological systems using permutated radar charts. PLoS ONE, 13.10.1371/journal.pone.0204309 ↗
54. An integrated framework of drivetrain degradation assessment and fault localization for offshore wind turbines
55. A new method for nondestructive quality evaluation of the resistance spot welding based on the radar chart method and the decision tree classifier10.1007/s00170-014-6654-1 ↗
56. Zagoruyko, S., and Komodakis, N. (2016). Wide residual networks. arXiv.10.5244/C.30.87 ↗
57. Object detection and location of robot based on Faster R-CNN
58. Classification of breast cancer histopathological images based on ResNet50 Network
59. Automatic segmentation for cell images bases on support vector machine and ellipse fitting
60. Chen, Y.P., Fan, H.Q., Xu, B., Yan, Z.C., Kalantidis, Y., Rohrbach, M., Yan, S.C., and Feng, J.S. (November, January 27). Drop an Octave: Reducing spatial redundancy in convolutional neural networks with Octave convolution. Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision, Seoul, Korea.10.1109/ICCV.2019.00353 ↗
61. Acoustic fault analysis of three commutator motors10.1016/j.ymssp.2019.07.007 ↗