Communication-based train control (CBTC) systems are automated train control systems based on continuous and bidirectional train-ground communication. As one of the critical subsystems of the safety-demanding CBTC systems, the Data Communication System (DCS) is physically exposed to the environment. DCS faults may lead to delayed trains, stranded passengers, suspension of service, or even catastrophic losses of lives or assets. Therefore, it is quite desirable to identify the DCS faults and accordingly take effective maintenance measures. Most fault classification works are based on traditional Machine Learning (ML) techniques, such as Decision Tree (DT) and Bayesian Network. A substantial amount of fault information can be lost when we manually extract features from the system logs. This paper proposes a Deep Hybrid Learning (DHL) method to identify the DCS faults. A pre-trained Chinese Bidi-rectional Encoder Representation from Transformers (BERT) Deep Learning (DL) model is used to extract fault features from raw text logs automatically, and several traditional ML models are used to classify the system faults by the extracted features. Our study highlights the advantages of the DHL model in DCS fault diagnosis tasks. The intense simulation results illustrate that the proposed DHL model achieves substantial accuracy compared with traditional ML algorithms.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Deep Hybrid Learning Based Fault Diagnosis for Data Communication Systems in Communication-Based Train Control Systems


    Beteiligte:
    Lin, Sen (Autor:in) / Zhu, Li (Autor:in) / Zhuang, Qingheng (Autor:in)


    Erscheinungsdatum :

    08.10.2022


    Format / Umfang :

    1166389 byte




    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Enhancing Communication-Based Train Control Systems Through Train-to-Train Communications

    Wang, Xiaoxuan / Liu, Lingjia / Tang, Tao et al. | IEEE | 2019



    Measurement-Based Handover Method for Communication-Based Train Control Systems

    Chan, Mei Ya / Baroudi, Sami / Siu, Joseph et al. | IEEE | 2017


    A Cognitive Control Approach to Communication-Based Train Control Systems

    Wang, Hongwei / Yu, F. Richard / Zhu, Li et al. | IEEE | 2015