A Warning Method of Traction Motor Filter Clogging Based onRecurrent Neural Network

Aiming at the problem of filter clogging in traction motor of urban rail transit vehicles, in order to reduce the difficulty and improve the instantaneity of after-sales service personnel in troubleshooting, a warning method of motor filter clogging based on recurrent neural network is proposed. The...

Full description

Saved in:
Bibliographic Details
Main Authors: ZHANG Shiqiang, DAI Jisheng, XU Hailong, ZHAN Yanhao
Format: Article
Language:zho
Published: Editorial Office of Control and Information Technology 2021-01-01
Series:Kongzhi Yu Xinxi Jishu
Subjects:
Online Access:http://ctet.csrzic.com/thesisDetails#10.13889/j.issn.2096-5427.2021.05.016
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Aiming at the problem of filter clogging in traction motor of urban rail transit vehicles, in order to reduce the difficulty and improve the instantaneity of after-sales service personnel in troubleshooting, a warning method of motor filter clogging based on recurrent neural network is proposed. The model is constructed based on the combination of motor principle and data-driven method, and actual operation data is used to train the model. By this method, motor temperature is predicted in real time based on the existing data of train, and motor filter clogging warning is realized according to the residual relationship between predicted values and actual values. Application results of a subway operation vehicle show that this method can accurately warn the clogging problem of motor filter, guide precise maintenance, and effectively promote the process of state maintenance.
ISSN:2096-5427