此页面上的内容需要较新版本的 Adobe Flash Player。

获取 Adobe Flash Player

Fault diagnosis method of track circuit based on KPCASAE


JIN Zuchen， DONG Yu


（School of Automation and Electrical Engineering, Lanzhou Jiaotong University, Lanzhou 730070, China）


Abstract： At present, ZPW2000 track circuit fault diagnosis is artificially analyzed and monitored. Its discrimination method not only is low efficient and takes a long period, but also requires highly experienced personnel to analyze the data. Therefore, we introduce kernel principal component analysis and stacked autoencoder network  (KPCASAD) into the fault diagnosis of ZPW2000 track circuit. According to the working principle and fault characteristics of track circuit, a fault diagnosis model of KPCASAE network is established. The relevant parameters of key components recorded in the data collected by  field staff are used as the fault feature parameters. The KPCA method is used to reduce the dimension and noise of fault document matrix to avoid information redundancy. The SAE network is trained by the processed fault data. The model parameters are optimized overall by using back propagation (BP) algorithm. The KPCASAE model is simulated in Matlab platform and is finally proved to be effective and feasible. Compared with the traditional method of artificially analyzing fault data and other intelligent algorithms, the KPCASAE based classifier has higher fault identification accuracy.


Key words： ZPW2000 track circuit; fault diagnosis; stacked autoencoder (SAE); kernel principal component analysis (KPCA)


References


［1］HUANG Z W, WEI X Y, LIU Z. Fault diagnosis of railway track circuit using fuzzy neural network. Journal of the China Railway Society, 2012, 34(11)：  5459.

［2］DONG Y, CHEN X. Intelligent fault diagnosis for ZPW2000 track circuit based on rough set theory and fuzzy cognitive map. Journal of the China Railway Society, 2018, 40(6)：  8389.

［3］HINTON G E, SIMON O, YEEWHYE T. A fast learning algorithm for deep belief nets. Neural Computation, 2006, 18(7)：  15271554.

［4］YANG T G, GUI W H. Fault diagnosis of induction motor based on KPCA and RVM. Journal of Motor and Control, 2016, 20(9)：  8995.

［5］WANG L H, XIE Y Y, ZHANG Y H. A fault diagnosis method for asynchronous motor using deep learning. Journal of Xi’an Jiaotong University, 2017, 51(10)：  128134. 

［6］LU C, WANG Z Y, QIN W L, et al. Fault diagnosis of rotary machinery components using a stacked denoising autoencoderbased health state identification. Signal Processing, 2016, 130：  377388. 

［7］LIN Y J. Railway signal basis. Beijing: China Railway Publishing House, 2009：  146148.

［8］China Railway. ZPW2000A noninsulated automatic blocking system. Beijing: China Railway Publishing House, 2013：  129.

［9］ZHOU Y B. Fault analysis and treatment of red light strip in ZPW2000A track circuit. Technology Innovation and Productivity, 2013(4)：  9899.

［10］BENGIO Y. Learning deep architectures for AI. Foundations and Trends in Machine Learning, 2009, 2(1)：  127.

［11］JU J B, HU S L, ZHU C, et al. A deep learning based method for equipment fault diagnosis. Electronics Optics and Control, 2018, 25(2)：  103106.

［12］VINCENT P, LAROCHELLE H, LAJOIE I, et al. Stacked denoising autoencoders：  learning useful representations in a deep network with a local denoising criterion. Journal of Machine Learning Research, 2010, 11(12)：  33713408.

［13］WEN J T, YAN C H, SUN J D, et al. Bearing fault diagnosis method based on compressed acquisition and deep learning. Chinese Journal of Scientific Instrument, 2018, 39(1)：  171179.

［14］HOU W Q, YE M, LI W H. Rolling element bearing fault classification using improved stacked denoising autoencoders. Journal of Mechanical Engineering, 2018, 54(7)：  8796.

［15］ZHANG C G, SONG J Z, JIANG J Q, et al. Imbalanced data classification algorithm of improved denoising autoencoder neural network. Application Research of Computers, 2017, 34(5)：  13291332.

［16］SALAKHUTDINOV R, HINTON G E. Replicated softmax：  an undirected topic model//The 23rd Annuel Conference on Neural Information Processing Systems, Dec. 710, 2009, Vancouver, BC, Canada. Massachusetts: Masachusetts Institute of Technology Press, 2009：  16071614.

［17］GUO Q J, YU H B, NIE Y Y, et al. Joint timefrequency and kernel principal component based SOM for machine Maintenance//The 3rd International Symposium on Neural Networks, May 28Jun.1, 2006, Chengdu, China. Berlin: Springer, 2006, 3973：  11441154.

［18］WANG Z Q, PANG X Q, HUANG R, et al. KPCAANFIS based remaining useful life prediction of lithium battery. Journal of Electronic Measurement and Instrumentation, 2008, 32(10)：  2632.

［19］QU X Y, ZENG P, XU C C, et al. Drop out denoising autoencoderbased fault diagnosis for grinding system. Control and Decision, 2018, 33(9)：  16621666.

［20］SHAN W P. Classification and recognition of transmission fault based on deep belief network. Guangzhou: South China University of Technology, 2015. 



基于KPCASAE网络的轨道电路故障诊断方法研究


金祖臣， 董昱


（兰州交通大学 自动化与电气工程学院， 甘肃 兰州 730070）


摘要:目前， ZPW2000型轨道电路故障判别主要依赖人工对数据进行分析， 该判别方式效率低、 周期较长、 对人工依赖程度高。 为此， 引入了栈式自编码网络(Stacked autoencoder, SAE)和核主元分析(Kernel principal component analysis, KPCA)的相关理论对ZPW2000型轨道电路进行故障诊断。 首先， 根据轨道电路的工作原理和故障特点， 建立了KPCASAE故障诊断模型。 然后， 将现场工作人员采集的数据中关键部件的相关参数作为故障特征参数， 利用KPCA对故障特征矩阵进行降维降噪处理， 避免信息冗余。 之后， 利用SAE网络对处理后的故障数据进行训练， 并采用BP算法对模型参数进行整体优化。 最后， 在Matlab环境下进行仿真实验。 仿真结果表明， 所提出的基于KPCASAE的轨道电路故障诊断模型是有效可行的， 与传统的故障数据人工分析方法及其他智能算法相比， 其故障识别准确率更高。 


关键词:ZPW2000型轨道电路； 故障诊断； 栈式自编码网络； 核主元分析


引用格式：JIN Zuchen， DONG Yu． Fault diagnosis method of track circuit based on KPCASAE． Journal of Measurement Science and Instrumentation， 2022， 13（1）： 8995. DOI： 10．3969／j．issn．16748042．2022．01．010


[full text view]