The invention discloses a federated learning-based traffic flow prediction method and system, and a medium. The method comprises the steps of setting hyper-parameters of a server model and initializing the hyper-parameters to obtain a global model; the server distributes the global model to the client to obtain each local model; each client updates the local model by using the local traffic flow data set; calculating the correlation between each local model update and the global model update, and screening clients according to the correlation; the screened clients send local model parameters to a server; the server gathers the received local model parameters to complete global model updating; repeating the steps until the models converge; and finally, each client predicts the traffic flow by using the converged local model. According to the method, invalid parameter uploading can be avoided, and the communication overhead in the federal learning training process is reduced.

    本发明公开了一种基于联邦学习的交通流预测方法、系统和介质,方法包括:设定服务器模型的超参数并初始化得到全局模型;服务器将全局模型分发到客户端得到各局部模型;各客户端利用本地交通流数据集更新局部模型;计算各局部模型更新与全局模型更新的相关性,根据相关性筛选客户端;筛选的客户端将其局部模型参数发送到服务器;服务器将接收到的局部模型参数进行汇聚,完成全局模型更新;重复上述步骤直至各模型收敛;最终各客户端使用收敛的本地局部模型对交通流进行预测。本发明可以避免无效参数上传,降低联邦学习训练过程的通信开销。


    Access

    Download


    Export, share and cite



    Title :

    Traffic flow prediction method and system based on federated learning, and medium


    Additional title:

    一种基于联邦学习的交通流预测方法、系统和介质


    Contributors:
    LU MINGMING (author) / HE WENYONG (author)

    Publication date :

    2023-04-04


    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 / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / H04L TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION , Übertragung digitaler Information, z.B. Telegrafieverkehr



    FedGRU: Privacy-preserving Traffic Flow Prediction via Federated Learning

    Liu, Yi / Zhang, Shuyu / Zhang, Chenhan et al. | IEEE | 2020


    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


    Traffic state scheduling method, system and device based on federated learning and medium

    YAN JUN / FENG SHU / WANG WEI | European Patent Office | 2024

    Free access