The invention relates to a ship traffic flow long sequence space-time prediction method based on ST-Informer. According to the ship traffic flow long sequence space-time prediction method, long sequence prediction is carried out on ship traffic flow by establishing a space-time fusion model. A method based on an ST-Informer model is mainly adopted for processing. The method comprises the following steps of: firstly, carrying out cleaning and preliminary statistical analysis on data of a ship automatic identification system (AIS), extracting ship traffic flow data of each harbor district in different time periods, secondly, carrying out default value filling and standardization processing on the ship traffic flow data, then, constructing an adjacent matrix and a characteristic matrix based on the distance between each harbor district and historical ship traffic flow, and finally, carrying out calculation on the adjacent matrix and the characteristic matrix. And predicting the ship traffic flow of each port based on an ST-Informer model. From the perspective of the topological structure and the long sequence of the water traffic road network, the time-space characteristics contained in the ship traffic flow data are fully considered, and the long sequence prediction precision of the ship traffic flow is effectively improved.
本发明涉及一种基于ST‑Informer的船舶交通流量长序列时空预测方法,本发明通过建立时空融合模型对船舶交通流量进行长序列预测。主要采用基于ST‑Informer模型的方法进行处理。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,提取各港区不同时间段的船舶交通流量数据,其次,对船舶交通流量数据进行缺省值填充和标准化处理,然后,基于各港区间距离和历史船舶交通流量构造邻接矩阵和特征矩阵,最后,基于ST‑Informer模型预测各港区船舶交通流量。本发明从水上交通路网的拓扑结构和长序列角度出发,充分考虑船舶交通流量数据中蕴含的时空特征,有效提高船舶交通流量长序列预测精度。
Ship traffic flow long sequence space-time prediction method based on ST-Informer
基于ST-Informer的船舶交通流量长序列时空预测方法
2023-09-29
Patent
Electronic Resource
Chinese
IPC: | G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / G07C TIME OR ATTENDANCE REGISTERS , Zeit- oder Anwesenheitskontrollgeräte / G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS |
Long-time-sequence traffic flow prediction method based on graph convolution-Informer model
European Patent Office | 2021
|Traffic flow prediction method based on space-time sequence
European Patent Office | 2025
|