Automatic Identification System (AIS) data mining is an important research direction in water transportation, and vessel trajectory prediction is one of its applications. Accurately predicting the future trajectory of a vessel is of great importance for situational awareness and risk warning in maritime traffic. However, the complexity of waterways, bank effect and navigational danger pose challenges to vessel trajectory prediction. To solve this problem, we proposed the Context-aware LSTM network(CaLSTM). This model is an encoder-decoder structure, which constructs channel information as an information matrix and uses convolutional neural network (CNN) and long short-term memory network (LSTM) to predict vessel trajectories. At the same time, a new loss function, enviro-MSE loss, was proposed to evaluate the environment of prediction results and reduce the possibility of prediction trajectory stranding. Finally, the proposed model was tested using data from Ningbo-Zhoushan port waters. The results show that the CaLSTM model can make more accurate and reasonable predictions, and the prediction results are more consistent with the ship navigation pattern than the baseline model under the complex environment.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Vessel trajectory prediction with the introduction of channel information


    Contributors:
    Lu, Xiaohu (author) / He, Murong (author) / Ma, Jie (author) / Wang, Han (author) / Shao, Yuanqi (author)


    Publication date :

    2023-08-04


    Size :

    1103614 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Vessel trajectory prediction in curving channel of inland river

    Xiaopeng, Tong / Xu, Chen / Lingzhi, Sang et al. | IEEE | 2015


    Vessel Trajectory Prediction Based on Context-Assisted Information

    Wang, Jianing / Jiao, Lianmeng / Pan, Quan | IEEE | 2024


    Short-term Inland Vessel Trajectory Prediction with Encoder-Decoder Models

    Donandt, Kathrin / Böttger, Karim / Söffker, Dirk | HENRY – Federal Waterways Engineering and Research Institute (BAW) | 2022

    Free access

    Short-term Inland Vessel Trajectory Prediction with Encoder-Decoder Models

    Donandt, Kathrin / Bottger, Karim / Soffker, Dirk | IEEE | 2022


    Vessel trajectory prediction with a Gated Spatio-Temporal Graph Aggregation Network

    Zhang, Xiliang / Liu, Jin / Gong, Peizhu et al. | IEEE | 2023