The invention relates to the technical field of train equipment, and concretely relates to a train traction motor temperature prediction method for establishing a neural network algorithm based on PyTorch. The method comprises the following steps: S1, carrying out data detection: selecting a plurality of trains, installing detection equipment on the trains so as to detect the daily temperature change of traction motors on the trains, and then summarizing and compiling information into a repository; and S2, carrying out data preprocessing: S2.1, performing data resampling: resampling the information acquired in the previous step, taking a temperature value every 60 points so as to obtain new data divided into time intervals, and sampling by adopting a nearest method so as to complement with the latest data of missing data. The method has the beneficial effects that one temperature value is taken every other time period, the pre-prediction time can be prolonged, the temperature change trend is relatively obvious, and the accuracy of data information can be improved by selecting and screening data according to the window characteristics and the time characteristics, so that the model prediction precision of the method is improved.

    本发明涉及列车设备技术领域,具体是一种基于PyTorch搭建神经网络算法的列车牵引电机温度预测方法,包括以下步骤:S1.数据检测:选取多个列车,并将检测设备安装在列车,从而检测列车上牵引电机日常温度变化,而后再将信息汇总编入储存库;S2.数据预处理:S2.1数据重采样:对上一步骤中采集的信息进行重采样,每隔60个点取一个温度值,从而得到一个以分为时间间隔的新数据,同时采用nearest方法进行采样,从而用缺失数据的最近一条数据进行补齐。本发明的有益效果每隔一个时间段取一个温度值,可以增大预测的前置时间,使得温度变化趋势较为明显,由窗口特征、时间特征选择筛分数据可以提高数据信息的准确度,从而提高该方法模型预测的精度。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Train traction motor temperature prediction method for establishing neural network algorithm based on PyTorch


    Weitere Titelangaben:

    一种基于PyTorch搭建神经网络算法的列车牵引电机温度预测方法


    Beteiligte:
    CHANG WEI (Autor:in) / YU JIEQUAN (Autor:in)

    Erscheinungsdatum :

    2021-09-10


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

    IPC:    G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / B61C Lokomotiven , LOCOMOTIVES



    Online Deep Learning for High-Speed Train Traction Motor Temperature Prediction

    Yang, Zhiqiang / Dong, Honghui / Man, Jie et al. | IEEE | 2024


    Traction motor of motor train unit

    YAN HAIYANG | Europäisches Patentamt | 2023

    Freier Zugriff

    Traction control method suitable for motor train unit train

    WAN ZHENG / LI XUEFEI / WANG ZHI et al. | Europäisches Patentamt | 2023

    Freier Zugriff

    PyTorch and automatic differentiation

    Meinert, Nis | Deutsches Zentrum für Luft- und Raumfahrt (DLR) | 2021

    Freier Zugriff

    MOTOR TRAIN UNIT EMERGENCY TRACTION SYSTEM

    DENG HAI / SHEN DI / CHEN TIANYU et al. | Europäisches Patentamt | 2020

    Freier Zugriff