基于机器学习的道岔故障诊断与预测研究综述  

Overview of Turnout Fault Diagnosis and Prognostic Based on Machine Learning

作　　者：谢博才 宫殿君 Xie Bocai;Gong Dianjun(China State Railway Group Co.,Ltd.,Beijing 100844,China;CRSC Research&Design Institute Group Co.,Ltd.,Beijing 100070,China;Beijing Engineering Technology Research Center of Operation Control Systems for High Speed Railways,Beijing 100070,China)

机构地区：[1]中国国家铁路集团有限公司,北京100844 [2]北京全路通信信号研究设计院集团有限公司,北京100070 [3]北京市高速铁路运行控制系统工程技术研究中心,北京100070

出　　处：《铁路通信信号工程技术》2021年第8期93-99,共7页Railway Signalling & Communication Engineering

基　　金：中国铁路总公司重大课题项目(2016G005-A)。

摘　　要：机器学习因其可以自适应的处理大量数据、实现智能分类和预测等优点,在道岔故障检测与诊断(Fault Detection and Diagnosis,FDD)技术方面已经有了许多理论研究和应用。对基于数据的道岔故障诊断算法进行分类和整理,重点介绍基于特征提取和基于数据的建模方法,对于机器学习算法的选择和运用具有一定的指导意义。最后通过分析道岔在预测与健康管理(Prognostics and Health Management,PHM)方面的研究现状,对机器学习在道岔故障预测与健康管理领域的应用进行展望。Machine learning has many theoretical researches and applications in turnout fault detection and diagnosis technology because of its advantages such as adaptive processing of large amounts of data,realization of intelligent classifi cation and prediction.This paper classifi es and sorts out data-based turnout fault diagnosis algorithms,and focuses on the feature extraction and data-based modeling methods,which have certain guiding signifi cance for the selection and application of machine learning algorithms.Finally,by analyzing the current research status of turnouts in prognostics and health management,the application of machine learning in the fi eld of turnout fault prognostics and health management is prospected.

关 键 词：道岔 机器学习 故障检测与诊断 预测与健康管理 

分 类 号：U213.6[交通运输工程—道路与铁道工程]