The invention discloses a distributed traffic flow prediction method based on personalized federal map signal learning, and the method comprises the steps: firstly calculating a significant period of each client through a periodogram method, uploading the significant period to a server, and enabling the server to cluster the clients with approximate periods; the method comprises the following steps of: firstly, calculating a characteristic value of a client graph structure by adopting a graph signal processing method, uploading the characteristic value and a model parameter to a server side, carrying out personalized aggregation operation on the server side by utilizing information uploaded by the client, generating a personalized prediction model for each client, issuing the personalized prediction model to the client, and generating a global model for the server side at the same time. According to the personalized aggregation operation designed by the invention, a cooperation mechanism between client traffic flow prediction models is constructed on the premise of not interacting traffic flow data between traffic areas, and the designed periodogram and the graph signal processing method enhance the effectiveness of cooperation by extracting the spatial-temporal characteristics of the client traffic flow data, so that the traffic flow prediction efficiency is improved. And the performance of the personalized model and the global model is improved.

    本发明公开了一种基于个性化联邦图信号学习的分布式交通流量预测方法,首先利用周期图法计算每个客户端的显著周期并上传至服务器端,服务器端将近似周期的客户端聚类成簇,然后采用图信号处理方法计算客户端图结构的特征值和模型参数一起上传至服务器端,再在服务器端利用客户端上传的信息进行个性化聚合操作,为每个客户端生成个性化预测模型并下发至客户端,同时为服务器端生成全局模型。本发明设计的个性化聚合操作,在不交互交通区域间流量数据的前提下构建客户端交通流量预测模型之间的协作机制,而设计的周期图和图信号处理方法,通过提取客户端交通流量数据的时空特征,增强协作的有效性,提升个性化模型和全局模型的性能。


    Access

    Download


    Export, share and cite



    Title :

    Distributed traffic flow prediction method based on personalized federal map signal learning


    Additional title:

    基于个性化联邦图信号学习的分布式交通流量预测方法


    Contributors:
    SHAN HANGGUAN (author) / YAN SIFU (author) / LI PU (author) / CHEN CHEN (author) / LI XINYU (author)

    Publication date :

    2024-09-20


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    Chinese


    Classification :

    IPC:    G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS / G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen



    Personalized federated learning traffic flow prediction method based on prior attention mechanism

    CHAI YIDONG / HUANG AMEI / LIU YI et al. | European Patent Office | 2024

    Free access

    Federal learning traffic prediction method based on traffic mode classification

    ZHANG XINMIN / SUN SHU / QIAN JINCHUAN et al. | European Patent Office | 2025

    Free access


    Traffic flow prediction method based on block chain and federal learning fusion

    ZHI HUI / DUAN MIAOMIAO / WANG YANING et al. | European Patent Office | 2023

    Free access

    Adaptive traffic signal control method combining traffic flow prediction and reinforcement learning

    PI JIATIAN / YANG XINMIN / WU CHANGZHI | European Patent Office | 2023

    Free access