Open Access Journal

ISSN : 2395-2717 (Online)

International Journal of Engineering Research in Computer Science and Engineering (IJERCSE)

Monthly Journal for Computer Science and Engineering

Open Access Journal

International Journal of Engineering Research in Electrical and Electronic Engineering(IJEREEE)

Monthly Journal for Electrical and Electronic Engineering

ISSN : 2395-2717 (Online)

Rectification of Fault Using Recurrent Neural Network Railway Track Circuit

Author : Mamatha G M 1 Bharath H P 2 Harshith H 3 Sowmiya C 4 Shivkumar Rathod 5

Date of Publication :30th November 2017

Abstract: Timely detection and identification of faults in railway track circuits are decisive for the safety and availability of railway networks. In this paper, the custom of the long-short-term memory (LSTM) recurrent neural network is proposed to undertake these tasks based on the commonly available measurement signals. By considering the signals from multiple track circuits in a geographic area, faults are spotted from their spatial and temporal dependences. A generative model is used to show that the LSTM network can learn these dependencies directly from the data. The network correctly classifies 99.7% of the test input sequences, with no false positive fault detections. In addition, the t-Distributed Stochastic Neighbour embedding (t-SNE) method is used to scrutinize the resulting network, further showing that it has learned the relevant dependencies in the data. Finally, we compare our LSTM network with a convolutional network trained on the same task. From this comparison, we conclude that the LSTM network architecture is better suited for the railway track circuit fault detection and identification tasks than the convolutional network.

Reference :

    1. J. Chen, C. Roberts, and P. Weston, ÒFault detection and diagnosis for railway track circuits using neuro-fuzzy systems,Ó Control Eng. Pract., vol. 16, no. 5, pp. 585Ð596, 2008.
    2. K. Verbert, B. De Schutter, and R. Babuška, ÒExploiting spatial and temporal dependencies to enhance fault diagnosis: Application to railway track circuits,Ó in Proc. Eur. Control Conf., Linz, Austria, Jul. 2015, pp. 3047Ð3052.
    3. S. Hochreiter and J. Schmidhuber, ÒLong short-term memory,Ó Neural Comput., vol. 9, no. 8, pp. 1735Ð1780, 1997
    4. R. Wu, S. Yan, Y. Shan, Q. Dang, and G. Sun. (2015). ÒDeep image: Scaling up image recognition.Ó [Online]. Available: http://arxiv.org/ abs/1501.02876
    5. A. Y. Hannun et al. (Dec. 2014). ÒDeep speech: Scaling up end-to-end speech recognition.Ó [Online]. Available: http://arxiv.org/ abs/1412.5567

Recent Article