As global economies continue to expand, maritime trade is projected to double by the year 2050. In this context, the development of a new generation of maritime intelligent supervision systems, which rely on precise information perception and machine learning-based technologies, is crucial for ensuring the safety of maritime navigation. Traditional methods for detecting abnormal ship behavior often suffer from inadequate feature extraction capabilities, leading to ineffective and inaccurate anomaly detection. To address this issue, this paper introduces a Transformer-GSA encoder detection model. This model enhances ship information data through BCE-GAN, extracts multi-layer convolutional features, and utilizes the Transformer-GSA module to capture the dependencies between trajectory features and time-series data, thereby effectively identifying abnormal ship behaviors. An evaluation conducted using AIS data from the Yantai-Qingdao route in January 2021 demonstrates that the model achieves an accuracy of 96.26%, with a reduced parameter count of 12.11M. These results confirm the model’s efficacy and superiority in detecting abnormal ship trajectories, contributing to the reduction of maritime accidents and illegal activities, alleviating the management burden on duty personnel, and enhancing the quality of intelligent ship traffic services and port operation efficiency under the maritime Internet of Things architecture.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Research on Ship Anomaly Detection Algorithm Based on Transformer-GSA Encoder


    Contributors:
    Liu, Hongdan (author) / Wu, Congjie (author) / Li, Bing (author) / Zong, Zhengcong (author) / Shu, Yaqing (author)


    Publication date :

    2025-06-01


    Size :

    2257742 byte




    Type of media :

    Article (Journal)


    Type of material :

    Electronic Resource


    Language :

    English




    An Auto-Encoder Based TinyML Approach for Real-Time Anomaly Detection

    Sai Charan, Kovuru | British Library Conference Proceedings | 2022


    An Auto-Encoder Based TinyML Approach for Real-Time Anomaly Detection

    Sai Charan, Kovuru | SAE Technical Papers | 2022


    Traffic anomaly detection method based on graph convolutional neural network auto-encoder

    LI XIAOJIE / REN ZHIYU / SHI CANGHONG et al. | European Patent Office | 2023

    Free access

    Spacecraft Anomaly Detection via Transformer Reconstruction Error

    Meng, Hengyu / Zhang, Yuxuan / Li, Yuanxiang et al. | Springer Verlag | 2020