Several elements influence how we manage the railway network, the most important of which are operational excellence and passenger satisfaction. Because traditional systems are unable to deal with the complexity and non-linearity of railway traffic data, most frequently use algorithmic approaches and simplistic statistical models. It cannot be used for better resource allocation; there will always be congestion and delays. To accomplish that, the paper proposes a novel approach that makes use of deep learning-based traffic prediction models such as Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM). The proposed method uses both real-time and historical traffic data, which improves prediction accuracy and operating efficiency. The model builds on existing methods by using CNNs to learn spatial patterns and LSTMs to understand time-series relationships. Other results demonstrate a significant improvement over bigger mistakes in existing systems (MAE = 0.75 and RMSE = 1.20). Operational indicators also reveal advantages, with 15% lower costs and optimal resource utilization. The CNN-LSTM model resulted in significant performance advantages, including a 25% reduction in fuel consumption and more than 30% reduction in maintenance expenses. The results show that bottleneck detection has the potential to improve precision and productivity over previous techniques, which could revolutionize railway traffic regulation.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Enhancing Railway Network Efficiency with Deep Learning-Based Traffic Prediction Models


    Weitere Titelangaben:

    Communic.Comp.Inf.Science


    Beteiligte:
    Rajagopal, Sridaran (Herausgeber:in) / Popat, Kalpesh (Herausgeber:in) / Meva, Divyakant (Herausgeber:in) / Bajeja, Sunil (Herausgeber:in) / Mudholkar, Pankaj (Herausgeber:in) / Radhika, C. (Autor:in) / Kerana Hanirex, D. (Autor:in)

    Kongress:

    International Conference on Advancements in Smart Computing and Information Security ; 2024 ; Rajkot, India October 16, 2024 - October 18, 2024



    Erscheinungsdatum :

    23.04.2025


    Format / Umfang :

    14 pages





    Medientyp :

    Aufsatz/Kapitel (Buch)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch




    Enhancing Global Maritime Traffic Network Forecasting with Gravity-Inspired Deep Learning Models

    Song, Ruixin / Spadon, Gabriel / Pelot, Ronald et al. | ArXiv | 2024

    Freier Zugriff

    Enhancing Traffic Efficiency through Real-Time Traffic Signal Control Using Deep Q-Learning

    Katragadda, Harshith / B, Natarajan / Bhuvaneswari, R | IEEE | 2023


    Short-Term Traffic Flow Prediction Based on Deep Learning Models

    Yuan, Bo / Li, Wanda / Li, Lin et al. | IEEE | 2024


    Traffic prediction with deep learning for vehicular communication network

    Wang, Jin / Liu, Peixin | British Library Conference Proceedings | 2022


    Urban Railway Network Traffic Prediction with Spatiotemporal Correlations Matrix

    Shao, Weijuan / Li, Man | British Library Conference Proceedings | 2016