The invention relates to a ship traffic flow prediction method based on an improved space-time diagram attention neural network, which predicts the ship traffic flow by establishing a space-time fusion model and mainly adopts a method based on the improved space-time diagram attention neural network for processing. The method comprises the following steps: firstly, carrying out cleaning and preliminary statistical analysis on data of an automatic ship identification system (AIS), extracting ship traffic flow data of each harbor district, then introducing a graph attention neural network (GAT) to improve a gated loop network (GRU) so as to identify the space-time relevance among the ship traffic flows of multiple harbor districts, and adding a time pattern attention mechanism (TPA) so as to identify the space-time relevance among the ship traffic flows of multiple harbor districts. And the internal relation of multiple harbor area time sequences is deeply mined, so that the modeling capability of the model on the time dependency relationship is improved. According to the method, from the perspective of the topological graph space of the water traffic road network, complex spatial and temporal characteristics in the marine traffic flow are comprehensively considered, and the ship traffic flow prediction effect can be effectively improved.

    本发明涉及基于改进时空图注意力神经网络的船舶交通流量预测方法,本发明通过建立时空融合模型对船舶交通流量进行预测,主要采用基于改进的时空图注意力神经网络方法进行处理。首先,对船舶自动识别系统(AIS)数据进行清洗和初步统计分析,提取各港区的船舶交通流量数据,然后引入图注意力神经网络(GAT)对门控循环网络(GRU)进行改进,以识别多港区船舶交通流量之间的时空关联性,并加入时间模式注意力机制(TPA),深度挖掘多个港区时间序列的内在联系,从而提高模型对时间依赖关系的建模能力。本发明从水上交通路网的拓扑图空间角度出发,综合考虑海上交通流量中复杂的时空特征,能够有效提高船舶交通流量预测效果。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Ship traffic flow prediction method based on improved space-time diagram attention neural network


    Weitere Titelangaben:

    基于改进时空图注意力神经网络的船舶交通流量预测方法


    Beteiligte:
    JIANG BAODE (Autor:in) / LUO HAIYAN (Autor:in) / SONG YUWEI (Autor:in)

    Erscheinungsdatum :

    2023-08-18


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

    IPC:    G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen



    Traffic flow speed prediction method based on attention space-time diagram convolutional network

    SUN YONG / ZHANG ANQIN / CHEN JINGJING | Europäisches Patentamt | 2023

    Freier Zugriff

    Traffic flow prediction method of attention-based deep residual space-time diagram convolutional network

    XU SHIJIAN / ZHANG XUHONG / WU YUEDONG et al. | Europäisches Patentamt | 2023

    Freier Zugriff

    Attention-based dynamic space-time diagram convolution traffic flow prediction system and method

    JI ZHENYUAN / HUAN YUEHUI / YANG GENKE et al. | Europäisches Patentamt | 2023

    Freier Zugriff

    Dynamic space-time neural network traffic flow prediction method based on attention mechanism

    MENG XIANGFU / XU RUIHANG / FAN HONGYU et al. | Europäisches Patentamt | 2023

    Freier Zugriff

    Traffic flow prediction method based on space-time diagram wavelet convolutional neural network

    MAO GUOJUN / ZHAO SHIHAO / WANG XIANG | Europäisches Patentamt | 2022

    Freier Zugriff