PSI - Issue 17

Takanori Hasegawa et al. / Procedia Structural Integrity 17 (2019) 487–494 Takanori Hasegawa et al./ Structural Integrity Procedia 00 (2019) 000 – 000

8

494

Table 4. Anomaly detection performance for DNN-based and AE-based fault detector.

DNN-based normal-anomaly classifier

AE-based novelty detector

operational conditions

normal 18944

faulty 1328 17613 7115 0.776

precision

normal 16470

faulty 3792 18422 7240 0.806

precision

normal

0.934 0.716 0.983

0.813 0.748

no-helium

7003

6194

bad-oil

125

0

1.0

recall

0.727

0.727

5. Conclusion

In the present study, data-driven approaches were applied to FESD condition monitoring. A CNN-based classifier was introduced to identify fault modes such as non-helium and bad-oil operation states and has shown to be effective in identifying bad-oil states. The inputs in no-helium states were challenging to identify because a scattering of data differs depending on recordings. More general anomaly detection was also examined. DNN-based normal-anomaly classifier and AE-based novelty detectors were developed on data-driven features extracted by the CNN; These systems yielded comparable performances and confused the no-helium state with the normal state in the same manner. These results indicate that mismatch between training and testing data cannot be negligible, and more data are needed during training.

Acknowledgement

The authors would like to thank Sanken Electric Co., Ltd. for providing a FESD and for useful discussions.

References

Akinyele, D. O., Rayudu, R. K., 2014. Review of energy storage technologies for sustainable power networks. Sustainable Energy Technologies and Assessments 8, 74 – 91. Amiryar, M. E., Pullen, K. R., 2017. A review of flywheel energy storage system technologies and their applications. Applied Sciences, 7(3), 286. Awata, A., Mita, A., 2006. Health monitoring of ball bearings in a flywheel using amplitude modulation of ultrasonic waves. Transactions of the Japan Society of Mechanical Engineers Series C 72(715), 729 – 734. Doucette, R. T., McCulloch, M. D., 2011. A comparison of high-speed flywheels, batteries, and ultracapacitors on the bases of cost and fuel economy as the energy storage system in a fuel cell based hybrid electric vehicle. Journal of Power Sources 196(3), 1163 – 1170. Hinton, G. E., Srivastava, N., Krizhevsky, A., Sutskever, I., Salakhutdinov, R., 2012. Improving neural networks by preventing co-adaptation of feature detectors. arXiv preprint arXiv: 1207.0580. Janssens, O., Slavkovikj, V., Vervisch, B., Stockman, K., Loccufier, M., Verstockt, S., Van de Walle, R., Van Hoecke, S., 2016. Convolutional neural network based fault detection for rotating machinery. Journal of Sound and Vibration 377(1), 331-345. Kateris, D., Moshou, D., Pantazi, X. E., Gravalos, I., Sawalhi, N., Loutridis, S., 2014. A machine learning approach for the condition monitoring of rotating machinery. Journal of Mechanical Science and Technology 28(1), 61 – 71. Kato, K., Ito, Y., Ishiguma, S., Nagano, T., Koiwa, K., Itoh, J., 2015. Ultra long lifetime energy storage system using flywheels and matrix converters, International Exhibition and Conference for Power Electronics, Intelligent Motion, Renewable Energy and Energy Management, Nuremberg, Germany. Krizhevsky, A., Sutskever, I., Hinton, G. E., 2012. Imagenet classification with deep convolutional neural networks, in Advances in neural information processing systems, 1097 – 1105. Liu, M., Xia, H., Sun, L., Li, B., Yang, Y., 2015. Vibration signal analysis of main coolant pump flywheel based on Hilberte-Huang transform. Nuclear Engineering Technology 47(2), 219 – 225. Nasiri, A., Hamidi, S. A., 2018. Uninterruptible power supplies. Power Electronics Handbook, 641 – 657. Tanabe, T., Suzuki, S., Ueda, Y., Ito, T., Numata, S., Shimoda, E., Funabashi, T., Yokoyama, R., 2009. Control performance verification of power system stabilizer with an EDLC in Islanded Microgrid. IEEJ Transactions on Power and Energy 129(1). Wicki, S., Hansen, E. G., 2017. Clean energy storage technology in the making: An innovation systems perspective on flywheel energy storage. Journal of Cleaner Production 162, 1118 – 1134.