High track quality is the basis of ensuring the safe and comfortable operation of trains, and it is of great significance to accurately grasp the degradation law of track geometry to ensure track quality. Long Short-Term Memory (LSTM) network in machine learning has the function of remembering historical information, which can better predict the development trend of nonlinear time series, a prediction model of orbital geometric degradation based on LSTM is constructed by using historical data of track quality index (TQI). In order to verify the validity of the model, the test data of Lanzhou-Urumqi High-speed Railway Track Inspection Vehicle are selected. The results show that the prediction model of track geometric degradation based on LSTM established in this paper is effective and has high prediction accuracy.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Prediction for geometric degradation of track on machine learning


    Contributors:
    Easa, Said (editor) / Wei, Wei (editor) / Hao, Junjie (author) / Liu, Rengkui (author)

    Conference:

    Eighth International Conference on Electromechanical Control Technology and Transportation (ICECTT 2023) ; 2023 ; Hangzhou, China


    Published in:

    Proc. SPIE ; 12790 ; 127903U


    Publication date :

    2023-09-07





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Prediction for geometric degradation of track on machine learning

    Hao, Junjie / Liu, Rengkui | British Library Conference Proceedings | 2023


    Railroad Track Geometric Degradation Analysis: A BNSF Case Study

    Rahimikelarijani, Behnam / Mohassel, Ahmad / Hamidi, Maryam | ASCE | 2019



    Fuzzy Approach in Rail Track Degradation Prediction

    Mostafa Karimpour / Lalith Hitihamillage / Najwa Elkhoury et al. | DOAJ | 2018

    Free access

    AN INTEGRATED MODEL FOR TRACK DEGRADATION PREDICTION

    Zhang, Y.-J. / Murray, M. / Ferreira, L. et al. | British Library Conference Proceedings | 1999