Advertisement

A Novel Unsupervised Time Series Discord Detection Algorithm in Aircraft Engine Gearbox

  • Zhongyu Wang
  • Dechang PiEmail author
  • Ya Gao
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11323)

Abstract

Aircraft engine discord detection is an important way to ensure flight safety. Unsupervised algorithms will be relatively effective due to the lack of models and tagged discord data. For the aircraft time series data collected from sensors, this paper proposes a Trend Featured Dynamic Time Wrapping for J Distance Discord Discovery algorithm based on the J-Distance Discord anomaly definition which combined with the trend information of the data. The experiments on aircraft engine gearbox data show that the TFDTW for JDD Discovery algorithm is better than the normal J-Distance Discord Discovery algorithm and also better than some other classic time series data discord detection algorithms.

Keywords

Time series data Unsupervised learning Abnormal detection Aircraft engine 

Notes

Acknowledgements

The research work is supported by National Natural Science Foundation of China (U1433116), the Fundamental Research Funds for the Central Universities (NP2017208) and Fundation of Graduate Innovation Center in NUAA (kfjj20171603).

References

  1. 1.
    Wang, T., Yu, J., Siegel, D., et al.: A similarity-based prognostics approach for remaining useful life estimation of engineered systems. In: International Conference on Prognostics and Health Management, PHM 2008, pp. 1–6. IEEE (2008)Google Scholar
  2. 2.
    Ryan, J., Lin, M.J., Miikkulainen, R.: Intrusion detection with neural networks. In: Advances in neural information processing systems, pp. 943–949. Morgan Kaufmann Publishers (1998)Google Scholar
  3. 3.
    Mukkamala, S., Janoski, G., Sung, A.: Intrusion detection using neural networks and support vector machines. In: Proceedings of the 2002 International Joint Conference on Neural Networks, IJCNN 2002, vol. 2, pp. 1702–1707. IEEE (2002)Google Scholar
  4. 4.
    Friedman, N., Geiger, D., Goldszmidt, M.: Bayesian network classifiers. Mach. Learn. 29(2–3), 131–163 (1997)CrossRefGoogle Scholar
  5. 5.
    Keogh, E., Lin, J., Lee, S.H., Van Herle, H.: Finding the most unusual time series subsequence: algorithms and applications. Knowl. Inf. Syst. 11(1), 127 (2007)Google Scholar
  6. 6.
    Keogh, E., Lin, J., Fu, A.: Hot SAX: efficiently finding the most unusual time series subsequence. In: Fifth IEEE International Conference on Data Mining, pp. 8. IEEE, November 2005Google Scholar
  7. 7.
    Huang, T., Zhu, Y., Wu, Y., et al.: J-distance discord: an improved time series discord definition and discovery method. In: IEEE International Conference on Data Mining Workshop, pp. 303–310. IEEE (2015)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2018

Authors and Affiliations

  1. 1.College of Computer Science and TechnologyNanjing University of Aeronautics and AstronauticsNanjingPeople’s Republic of China

Personalised recommendations