Accurate short-term four-dimensional trajectory prediction (TP) can enhance conflict detection capability and facilitate informed decision making for conflict resolution. The challenge of trajectory prediction lies in considerable uncertainties, especially the uncertainty introduced by weather effects. To address this challenge, we employ the Long Short-Term Memory (LSTM) neural network, renowned for its ability to forecast future time series. By harnessing a combination of historical trajectory data and weather data, our implementation seeks to predict the trajectory in the immediate future. A Bayesian Neural Network (BNN) is integrated to address the inherent uncertaintiy in the model, allowing for more robust and reliable predictions. The robustness and accuracy of the proposed method and model are rigorously validated using national meteorological data and ADS-B data. This validation procedure serves to thoroughly assess the performance of the model across various scenarios and ascertain its ability to generate precise predictions.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Short-term Aircraft Trajectory Prediction Considering Weather Effect


    Beteiligte:
    Feng, Shuai (Autor:in) / Wang, Gang (Autor:in) / Zhao, Peng (Autor:in) / Chao, Xu (Autor:in) / Cai, Kaiquan (Autor:in)


    Erscheinungsdatum :

    2023-10-01


    Format / Umfang :

    3173688 byte





    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch