Overview of Turnout Fault Diagnosis and Prognostic Based on Machine Learning

doi:10.3969/j.issn.1673-4440.2021.08.021

Railway Signalling & Communication Engineering ›› 2021, Vol. 18 ›› Issue (8): 93-99.DOI: 10.3969/j.issn.1673-4440.2021.08.021

Previous Articles Next Articles

Overview of Turnout Fault Diagnosis and Prognostic Based on Machine Learning

Xie Bocai¹, Gong Dianjun^2，3

1. China State Railway Group Co., Ltd., Beijing 100844, China；
2. CRSC Research & Design Institute Group Co., Ltd., Beijing 100070, China；
3. Beijing Engineering Technology Research Center of Operation Control Systems for High Speed Railways, Beijing 100070, China

Received:2021-06-02 Revised:2021-07-08 Online:2021-08-25 Published:2021-08-25

基于机器学习的道岔故障诊断与预测研究综述

谢博才¹，宫殿君^2，3

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

基金资助:
中国铁路总公司重大课题项目（2016G005-A）

Abstract

Abstract: 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 classification and prediction. This paper classifies and sorts out data-based turnout fault diagnosis algorithms, and focuses on the feature extraction and data-based modeling methods, which have certain guiding significance 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 field of turnout fault prognostics and health management is prospected.

Key words: turnout, machine learning, fault detection and diagnosis, prognostics and health management

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。

关键词: 道岔, 机器学习, 故障检测与诊断, 预测与健康管理

CLC Number:

U213.6

Xie Bocai, Gong Dianjun. Overview of Turnout Fault Diagnosis and Prognostic Based on Machine Learning[J]. Railway Signalling & Communication Engineering, 2021, 18(8): 93-99.

谢博才, 宫殿君. 基于机器学习的道岔故障诊断与预测研究综述[J]. 铁路通信信号工程技术, 2021, 18(8): 93-99.

Figures/Tables 2

References

[1]翟嘉琪，杨希祥，程玉强，等.机器学习在故障检测与诊断领域应用综述[J].计算机测量与控制，2021，29(3):1-9.
Zhai Jiaqi, Yang Xixiang, Cheng Yuqiang, et al. Overview of Application of Fault Detection and Diagnosis Based on Machine Learning[J]. Computer Measurement & Control, 2021, 29(3):1-9.
[2]郭一帆，唐家银.基于机器学习算法的寿命预测与故障诊断技术的发展综述[J].计算机测量与控制，2019，27(3):7-13.
Guo Yifan, Tang Jiayin. A Review of the Development of Life Prediction and Fault Diagnosis Technology Based on Machine Learning Algorithm[J]. Computer Measurement & Control, 2019, 27(3): 7-13.
[3]汪倩，周振威，陆裕东，等.转辙机的故障预测与健康管理技术[J].都市快轨交通，2015，28(1):58-61.
Wang Qian, Zhou Zhenwei, Lu Yudong, et al. Prognostics and Health Management Technology of Point Machine[J]. Urban Rapid Rail Transit, 2015, 28(1): 58-61.
[4]杨静.道岔故障诊断方法研究[D].南京:南京理工大学，2017.
[5]孙孟雷.基于神经网络的道岔转辙机故障诊断方法研究[D].成都:电子科技大学，2020.
[6] Yang Juhua, Li Xutong, Chen Guangwu, et al. Turnout Fault Diagnosis Based on DBSCAN/SOM[J]. Journal of Measurement Science and Instrumentation, 2021(12): 1-10.
[7]李孟澜.基于聚类集成的转辙机故障数据挖掘[D].成都:西南交通大学，2018.
[8]朱绪昌.基于灰关联的道岔信号控制系统故障诊断方法及其实现研究[D].兰州:兰州交通大学，2016.
[9] Wang Dengfei, Chen Guangwu, XingDongfeng, et al. Fault Diagnosis Method for Switch Control Circuit Based on SVM-AdaBoost[J]. Journal of Measurement Science and Instrumentation, 2020, 11(3): 251-257.
[10]孔令刚，焦相萌，陈光武，等.基于Mallat小波分解与改进GWO-SVM的道岔故障诊断[J].铁道科学与工程学报，2020，17(5):1070-1079.
Kong Linggang, Jiao Xiangmeng, Chen Guangwu, et al. Turnout Fault Diagnosis Based on Mallat Wavelet Decomposition and Improved GWO-SVM[J]. Journal of Railway Science and Engineering, 2020, 17(5): 1070-1079.
[11]习家宁.基于机器学习的集中监测道岔故障诊断研究与设计[D].北京:北京交通大学，2020.
[12]陈亭.基于道岔动作电流的故障诊断方法研究[D].成都:西南交通大学，2018.
[13]李林霄.基于改进卷积神经网络的道岔故障诊断方法研究[D].兰州:兰州交通大学，2019.
[14]杨菊花，于苡健，陈光武，等.基于CNN-GRU模型的道岔故障诊断算法研究[J].铁道学报，2020，42(7):102-109.
Yang Juhua, Yu Yijian, Chen Guangwu, et al. Research on Turnout Fault Diagnosis Algorithms Based on CNN-GRU Model[J]. Journal of the China Railway Society, 2020, 42(7): 102-109.
[15]左元.基于监测电流曲线的道岔故障诊断关键技术研究[D].西安:西安理工大学，2020.
[16]侯大山.基于神经网络的道岔故障预测和诊断[D].北京:北京交通大学，2019.
[17]钟志旺.铁路道岔健康状态评估与预测方法研究[D].北京:北京交通大学，2019.
[18]许庆阳.道岔故障诊断及健康状态预测[D].北京:北京交通大学，2018.
[19]高利民，许庆阳，李锋，等.基于SOM-BP混合神经网络的道岔设备退化状态研究[J].中国铁道科学，2020，41（3）：50-58.
Gao Limin, Xu Qingyang, Li Feng, et al. Research on Degradation State of Turnout Equipment Based on SOM-BP Hybrid Neural Network[J]. China Railway Science, 2020, 41(3): 50-58.
[20]李园祥.基于神经网络的道岔转辙机退化模型的研究[D].北京:北京交通大学，2020.

Overview of Turnout Fault Diagnosis and Prognostic Based on Machine Learning

基于机器学习的道岔故障诊断与预测研究综述

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 2

References

Related Articles 8

Recommended Articles

Metrics

Comments

[1]	Ning Yongmei. Discussion on Technical Scheme of Balise Group of Continuous Large Number Turnout [J]. Railway Signalling & Communication Engineering, 2021, 18(8): 15-18.
[2]	Xu Guifu. Train Control Scheme Selection in Scenario of Short Connecting Line Overlapped with Continuous Large-number Turnout [J]. Railway Signalling & Communication Engineering, 2021, 18(8): 19-23.
[3]	Wu Haiping. Dealing with Mixed Circuit Fault of ZD6 Turnout with Current Method [J]. Railway Signalling & Communication Engineering, 2021, 18(8): 69-74.
[4]	Wang Jianqiang, Lu Jianfeng, Chen Zhihui, Yang Jizhong. Discussion on Conversion Scheme of Movable Rack in Rack Rail Turnout Section [J]. Railway Signalling & Communication Engineering, 2021, 18(7): 22-24,53.
[5]	Gao Xinjun, Wang Wentao, Zhou Yongjian, He Leiqiang, Niu Qin. Research on Test Scene of Large-number Turnout for High Speed Railway Train Control System [J]. Railway Signalling & Communication Engineering, 2021, 18(4): 1-4.
[6]	Ning Dijie. Research on Turnout Auxiliary Protection System in Straddle-monorail System [J]. Railway Signalling & Communication Engineering, 2021, 18(2): 84-87.
[7]	Chen Zetao, Zheng Lefan. Improving Turnout Starting Circuit of ZYJ7-type Point Machine [J]. Railway Signalling & Communication Engineering, 2021, 18(2): 104-107.
[8]	Zhang Changping, Zhou Zhanghai. Impact and Analysis of Signaling Turnout Design Scheme on Interlocking Relation [J]. Railway Signalling & Communication Engineering, 2021, 18(1): 91-94.