Intelligent Fault Detection of High-Speed Railway Turnout Based on Hybrid Deep Learning

  • Zhi Zhuang
  • Guohua ZhangEmail author
  • Wei Dong
  • Xinya Sun
  • Chuanjiang Wang
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11320)


With the purpose of detecting the turnout fault without label data and fault data timely, this paper proposes a hybrid deep learning framework com-bining the DDAE (Deep Denoising Auto-encoder) and one-class SVM (Support Vector Machine) for turnout fault detection only using normal data. The proposed method achieves an accuracy of 98.67% on the real turn-out dataset for current curve, which suggests that this work realizes the purpose of detecting the fault with only normal data and provides a basis for the intelligent fault detection of turnouts.


Fault detection Deep Denoising Auto-encoder DBSCAN One-class SVM 



This work was supported by the National Key Research and Development Program of China under Grant 2017YFB1200700, the special fund of Suzhou-Tsinghua Innovation Leading Action under Grant 2016SZ0202, the Natural Science Foundation of China under Grants 61490701 and the Research and Development Project of Beijing National Railway Research & Design Institute of Signal & Communication Ltd.


  1. 1.
    Feng, L.: Turnout fault diagnosie by current analysis using computer monitor. Technol. Exch. 8(1), 73–75 (2011)Google Scholar
  2. 2.
    Huang, B.: Fault analysis and countermeasures for turnout switching equipment of high-speed railway. Railw. Oper. Technol. 22(1), 59–64 (2016)Google Scholar
  3. 3.
    Liu, X., Kuang, W., He, T.: The design of point switch simulator of high-speed turnout diagnostic apparatus. Railw. Stand. Des. 59(12), 105–110 (2015)Google Scholar
  4. 4.
    Zhang, Y., Xie, Q.: A method of railway turnout detection based on machine vision. Comput. Appl. Softw. 32(1), 225–228 (2015)Google Scholar
  5. 5.
    Bocaniala, C.D., Costa, J.S.D.: Application of a novel fuzzy classifier to fault detection and isolation of the DAMADICS benchmark problem. Control. Eng. Pract. 14(6), 653–669 (2006)CrossRefGoogle Scholar
  6. 6.
    Eker, O.F., Camci, F., Guclu, A., Yilboga, H., Sevkli, M., Baskan, S.: A simple state-based prognostic model for railway turnout systems. IEEE Trans. Industr. Electron. 58, 1718–1726 (2011)CrossRefGoogle Scholar
  7. 7.
    Zhang, K., Du, K., Ju, Y.: Algorithm of railway turnout fault detection based on PNN neural network. In: Seventh International Symposium on Computational Intelligence and Design, pp. 544–547. IEEE Computer Society (2014)Google Scholar
  8. 8.
    Zhang, K.: The railway turnout fault diagnosis algorithm based on BP neural network. In: IEEE International Conference on Control Science and Systems Engineering, pp. 135–138 (2015)Google Scholar
  9. 9.
    Zhou, F., Xia, L., Dong, W., et al.: Fault diagnosis of high-speed railway turnout based on support vector machine. In: IEEE International Conference on Industrial Technology, pp. 1539–1544. IEEE (2016)Google Scholar
  10. 10.
    He, Y.M., Zhao, H.B., Tian, J., Zhang, M.Q.: Railway turnout fault diagnosis based on support vector machine. Appl. Mech. Mater. 556-562, 2663–2667 (2014)CrossRefGoogle Scholar
  11. 11.
    Eker, O.F., Camci, F., Kumar, U.: SVM based diagnostics on railway turnouts. Int. J. Perform. Eng. 8(8), 289–298 (2012)Google Scholar
  12. 12.
    Wang, S.M., Lei, Y.: Fault diagnosis of turnout control circuit based on LS-SVM. J. Lanzhou Jiaotong Univ. 29(4), 1–5 (2010)MathSciNetGoogle Scholar
  13. 13.
    Vileiniskis, M., Remenyte-Prescott, R., Rama, D.: A fault detection method for railway point systems. Proc. Inst. Mech. Eng. Part F J. Rail Rapid Transit 230(3), 18 (2016)CrossRefGoogle Scholar
  14. 14.
    Shan, J.: Study on railway switches diagnosis system based on deep belief networks. Master’s thesis, Shijiazhuang Tiedao University (2017). (in Chinese)Google Scholar
  15. 15.
    Wu, Z., Wang, X., Jiang, Y.G., Ye, H., Xue, X.: Modeling spatial-temporal clues in a hybrid deep learning framework for video classification. In: ACM International Conference on Multimedia, pp. 461–470 (2015)Google Scholar
  16. 16.
    Du, S., Li, T., Gong, X.: Traffic flow forecasting based on hybrid deep learning framework. In: International Conference on Intelligent Systems and Knowledge Engineering, pp. 1–6 (2018)Google Scholar
  17. 17.
    Wang, H., Dong, W., Ye, H., Yan, Y., Yan, X.: Clustering of S700K point machine’s current curves based on reducing dimensions with denoising autoencoders and t-SNE. In: Chinese Automation Congress and Intelligent Manufacturing International Conference, pp. 742–747 (2017). (in Chinese)Google Scholar
  18. 18.
    Wuhan Railway Bureau: A Guide for Information Analysis of Signal Concentration Detection. China Railway Publishing House, Beijing (2015). (in Chinese)Google Scholar
  19. 19.
    Tax, D.M.J., Duin, R.P.W.: Support vector domain description. Pattern Recognit. Lett. 20(11–13), 1191–1199 (1999)CrossRefGoogle Scholar
  20. 20.
    Akhtar, S., Kumar, A., Ekbal, A., Bhattacharya, P.: A hybrid deep learning architecture for sentiment analysis. In: COLING (2016)Google Scholar
  21. 21.
    Sun, Y.A., Mao, H., Sang, Y.S., Zhang, Y.: Explicit guiding auto-encoders for learning meaningful representation. Neural Comput. Appl. 28(3), 1–8 (2015)Google Scholar
  22. 22.
    Sun, Y.A., Xue, B., Zhang, M.J., Yen, G.G.: An experimental study on hyper-parameter optimization for stacked auto-encoders. In: IEEE Congress on Evolutionary Computation (2018, in press)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2018

Authors and Affiliations

  • Zhi Zhuang
    • 1
    • 2
  • Guohua Zhang
    • 2
    Email author
  • Wei Dong
    • 2
  • Xinya Sun
    • 2
    • 3
  • Chuanjiang Wang
    • 1
  1. 1.Shandong University of Science and TechnologyQingdaoChina
  2. 2.Beijing National Research Center for Information Science and Technology (BNRist)Tsinghua UniversityBeijingChina
  3. 3.Department of AutomationTsinghua UniversityBeijingChina

Personalised recommendations