计算机科学

计算机科学 ›› 2024, Vol. 51 ›› Issue (9): 331-337.doi: 10.11896/jsjkx.231200190

• 计算机网络 • 上一篇    下一篇

面向轨道交通智能故障检测的边云计算方法

李志1,2, 林森1, 张强3   

  1. 1 北京交通大学自动化与智能学院 北京 100044
    2 中国电子科技集团公司第十五研究所 北京 100083
    3 中国人民解放军61741部队 北京 100070
  • 收稿日期:2023-12-27 修回日期:2024-05-24 出版日期:2024-09-15 发布日期:2024-09-10
  • 通讯作者: 林森(21120216@bjtu.edu.cn)
  • 作者简介:(22115005@bjtu.edu.cn)

Edge Cloud Computing Approach for Intelligent Fault Detection in Rail Transit

LI Zhi1,2, LIN Sen1, ZHANG Qiang3   

  1. 1 School of Automation and Intelligence,Beijing Jiaotong University,Beijing 100044,China
    2 The 15th Research Institute of China Electronics Technology Group Corporation,Beijing 100083,China
    3 The Chinese People's Liberation Army 61741 Unit,Beijing 100070,China
  • Received:2023-12-27 Revised:2024-05-24 Online:2024-09-15 Published:2024-09-10
  • About author:LIN Sen,born in 1999,postgraduate.His main research interests include urban rail transit and artificial intelligence.

PDF (PC)

329

摘要： 轨道交通系统是当今社会中交通运力的主要承载系统,对安全性有极高的要求。轨道交通系统的多个组件由于直接暴露在环境中,受多种外界因素影响,易出现故障。这些故障可能会导致列车延误、乘客滞留、服务暂停,甚至是灾难性的生命或资产损失。因此,需要设计针对轨道交通系统的实时故障检测方案,进而才能采取有效的维护措施。不同于基于传统的机器学习(Machine Learning,ML)的故障检测工作,本研究采用中文双向编码器表示转换器(Bidirectional Encoder Representation from Transformer,BERT)深度学习(Deep Learing,DL)模型进行实时的智能故障检测。该模型能够在处理故障检测任务时获取双向上下文的理解,从而更准确地捕捉句子中的语义关系,使得其对故障描述的理解更为精准。BERT的训练需要大量的数据支持,而轨道交通领域中存在多个运营商,它们各自持有独立的故障检测数据。由于数据的保密性,这些数据无法进行共享,从而限制了模型的训练,故采用了联邦边云计算方法,允许多个运营商在保持数据隐私的前提下共同训练BERT模型。联邦学习结合边云计算方法,在本地对轨道交通各运营商的数据进行初步处理,然后将汇总后的梯度上传至云端进行模型训练,最终将训练得到的模型参数发送回各边缘设备,实现模型的更新。研究结果表明,采用联邦边云计算方法进行BERT模型训练,在轨道交通领域的故障检测任务中优于目前已有的先进方案。这一方法在解决数据保密性问题的同时,有效提升了轨道交通故障检测的准确性与可靠性。

关键词: 轨道交通, 故障检测, 边云计算, 联邦学习, BERT

Abstract: Rail transit systems are the main carrying system of transportation capacity in the current society.It is extremely sensitive to safety.Because multiple components of the system are directly exposed to the environment,they are affected by various environments and are prone to failures,which may cause train delays,passenger retention,service outage,or even catastrophic loss of life or property.Therefore,it is necessary to design a fault detection scheme so that effective maintenance measures can be taken.Different from traditional machine learning(ML) based fault classification work,this paper adopts Chinese bidirectional encoder representation from transformer(BERT) deep learning(DL) model for intelligent fault detection.The model can obtain bidirectional contextual understanding when dealing with fault detection tasks,so as to more accurately capture the semantic relationship in sentences,and understand the fault descriptions more accurately.The training of BERT requires a large amount of data support,and there are multiple operators in the field of rail transit,each of which holds independent fault detection data.Due to the confidentiality of the data,these data cannot be shared,which limits the training of the BERT model.This paper designs and adopts the federated edge cloud computing method,allowing multiple operators to jointly train the BERT model while maintaining data privacy.Federated learning combined with the edge cloud computing method allows the data of rail transit operators to be preliminary processed locally,and then the summarized gradients are uploaded to the cloud for model training,and finally the trained model parameters are sent back to each edge device to realize model updates.The research results show that the BERT model training using the federated edge cloud computing method is superior to the existing advanced solutions in the fault detection task in the field of rail transit.This method not only solves the problem of data confidentiality,but also effectively improves the accuracy and reliability of fault detection.

Key words: Rail transit, Fault detection, Edge cloud computing, Federated learning, Bidirectional encoder representation from transformer

中图分类号: 

  • U285.4
[1]WEI X D.Urban rail transit automation system and technology[M].Electronic Industry Press,2004.
[2]JIANG F,Research on the connection between urban rail transit and other modes of transportation [J].Journal of North Jiaotong University,2001(4):108-110.
[3]TI Y.Development status and countermeasures of urban railtransit in China [J].Science,Technology and Economy of Inner Mongolia,2005(6):317-319.
[4]XU K,ZHANG C B,CHEN Z H.Overview of rail transit fault diagnosis [C]//The 18th Annual Conference of China System Simulation Technology and its Application.2003.
[5]GONG Y M,CHEN X K,ZHANG G F.Information security ofthe urban rail transit signaling system [J].Electronic Communication and Computer Science,2023,5(4):89-91.
[6]LIU T S,ZHU X J,XU R H.Analysis of the operation safetyand reliability of the urban rail transit system [J].Research on Urban Rail Transit,2006,9(1):15-17.
[7]SUN Y J,ZHANG S,MIAO C X,et al.Improved BP Neural Network forTransformer Fault Diagnosis[J].Journal of China University of Mining & Technology－English Edition,2007,17(1):138-142.
[8]CHEN K W,WEI S G,SHI H B,et al.An on-board network fault diagnosis method based on machine learning:CN202211088642.2[P].[2024-02-25].http://www2.drugfuture.com/cnpat/verify.aspx?cnpatentno=202211088642.2.
[9]ZHOU J B,XIAO M H,ZHU Y J, et al.Study on fault diagnosis of tractor diesel engine based on HPO-SVM [J].Journal of Nanjing Agricultural University,2023,46(2):12.
[10]WU T J.Data-driven fault diagnosis method based on machine learning [D].Hangzhou:Zhejiang University of Technology,2020.
[11]ZHANG J P,WANG L X.Simulation study on individualizeddiagnosis of circuit faults [J].Comkputer Simulation,2022(8):168-170.
[12]WANG Q,MEN X Z.Fault Diagnosis Based on Rough Sets and C4.5 dsDecision Tree[J].Journal of Northeastern University:Natural Science,2006,27(10):1138-1141.
[13]LIU D,ZHANG Y,ZOU G Y.Fault diagnosis method,device,equipment and medium based on decision tree:CN202211667676.7[P].[2023-04-27].http://www2.drugfuture.com/cnpat/verify.aspx?cnpatentno=202211667676.7.
[14]WANG X M,FU Y B,LI J T.Diagnosis diagnosis decisionmethod for power equipment based on decision tree algorithm [J].Automation Application,2023,64(12):128-131.
[15]ZHAO W,ZHU Y,WANG X.Combinatorial Bayes network in fault diagnosis of power transformer[J].Electric Power Automation Equipment,2009,29(11):6-9.
[16]WANG Z Y,SONG R W,SHI H.An online diagnosis method of bearing faults based on Dynamic Bayesian network [J].Journal of Taiyuan University of Science and Technology,2023,44(4):303-308.
[17]WANG Z H.Methods for ventilator fault diagnosis based on a Bayesian network [J].Equipment Management and Maintenance,2023(4):150-151.
[18]LIU K,YUAN Y Y.Feature extraction and clustering of short articles based on autoencoder [J].Journal of Peking University:Natural Science Edition,2015,51(2):282-288.
[19]DEVLIN J,CHAN M W,LEE K,et al.Bert:Pre-training of deep bidirectional transformers for language understanding[J].arXiv:1810.04802,2018.
[20]DU X X,LIU Y.Research on Beijing Rail Transit Information Exchange Platform based on XML [J].Railway Computer Applications,2012,21(1):55-61.
[21]YUE C,ZHANG K,QU S H.Application of urban information model in shield tunneling engineering [J].Urban Rail Transit research,2021,24(7):225-229.
[22]SUN M,FU Z.A method and device for predicting short-timepassenger flow of rail transit:CN202111203957.2 [P].[2024-02-25].http://www2.drugfuture.com/cnpat/verify.aspx?cnpatentno=202111203957.2.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发