SVDD-Based Research on Railway-Turnout Fault Detection and Health Assessment
-
摘要: 为更加精准地提高现场道岔维护质量，减少设备故障，提出了基于支持向量描述（support vector domain description，SVDD）的道岔转换故障检测和健康指标评估方法.首先，通过对道岔设备的动作机理分析，将道岔功率曲线分为3个动作过程：解锁过程，转换过程，锁闭过程，每一过程分别提取道岔功率曲线相应的特征参数；然后，在此特征空间上，设计了基于SVDD的道岔故障检测算法和健康评估算法，从而实现对道岔设备的健康管理；最后，利用道岔系统数学模型仿真数据和现场采集的道岔真实监测数据，对提出的算法有效性进行验证.实验表明，用本文提出的健康指标检测道岔故障和评估道岔健康状态的准确率达95%以上，对现场道岔维护有很大指导意义.Abstract: In order to improve the quality of railway-turnout maintenance procedures and reduce equipment failure, a method for turnout-fault-detection and health-index evaluation based on the support vector domain description (SVDD) technique, is proposed. Through analysis of the switch-equipment mechanism, the turnout power curve was divided into 3 parts, each representing the unlocking, conversion, and locking processes respectively. To this end, characteristic parameters of the switch power curve were extracted. Subsequently, in the feature space, an SVDD-based turnout-fault-detection algorithm and health assessment were proposed. Lastly, the efficiency of the proposed algorithm was verified by comparing simulation results against experimental data. Experimental results demonstrate that the proposed health index possesses an accuracy of 95%, which could prove to be significant in on-site maintenance and guidance applications.
-
Key words:
- railway turnouts /
- SVDD /
- fault detection /
- health evaluation
-
图 1 道岔转换设备电流曲线和功率曲线采集原理
Figure 1. Data-collection principle for turnout current and power
图 2 道岔转换过程中的典型功率曲线
Figure 2. Typical power curves observed during switch-conversion process
图 4 道岔在“尖轨摩擦力变大”故障模式演变下的电流曲线
Figure 4. Turnout current under fault mode of switch under increased friction
图 6 道岔动作过程中健康指标和故障检测阈值
Figure 6. Health-index and fault-detection thresholds during switch conversion
-
沈聪.道岔综合监测技术的研究[D].北京: 中国铁道科学研究院, 2016. http://cdmd.cnki.com.cn/Article/CDMD-83801-1016175026.htm 王军.高速铁路道岔智能监测系统的研发[D].兰州: 兰州交通大学, 2013. http://cdmd.cnki.com.cn/Article/CDMD-10732-1013352533.htm 钟志旺, 陈建译.基于CMOS面阵的道岔缺口定位监测研究[J].铁道学报, 2016, 38(12):70-75. doi: 10.3969/j.issn.1001-8360.2016.12.011ZHONG Zhiwang, CHEN Jianyi. CMOS plane based location and detection of switch gaps[J]. Journal of the China Railway Society, 2016, 38(12):70-75. doi: 10.3969/j.issn.1001-8360.2016.12.011 中国铁路总公司.铁路信号集中监测系统应用与维护技术[M].北京:中国铁道出版社, 2013:40-60. 许伟, 马建, 汪寿阳.系统分析方法集成研究及其在预测和监测中的应用[M].北京:科学出版社, 2011:20-95. 冯国瑜, 肖怀铁, 付强, 等.基于ISVDD的雷达高分辨距离像在线识别方法[J].系统工程与电子技术, 2012, 34(3):489-494. doi: 10.3969/j.issn.1001-506X.2012.03.11FENG Guoyu, XIAO Huaitie, FU Qiang, et al. Online recognition method of HTTP for radar systems based on ISVDD[J].System Engineering and Electronics, 2012, 34(3):489-494. doi: 10.3969/j.issn.1001-506X.2012.03.11 罗慧, 王友仁.基于GSM-SVDD的模拟电路故障诊断方法[J].电机与控制学报, 2013, 17(1):108-113. doi: 10.3969/j.issn.1007-449X.2013.01.018LUO Hui, WANG Youren. A GSM-SVDD method for analog circuits fault diagnosis[J]. Electric Machines And Control, 2013, 17(1):108-113. doi: 10.3969/j.issn.1007-449X.2013.01.018 张敏龙, 王涛, 王旭平, 等.带故障样本的弹性双阈值SVDD在线故障诊断算法及其应用[J].振动工程学报, 2016, 29(6):555-560. http://d.old.wanfangdata.com.cn/Periodical/zdgcxb201603023ZHANG Minlong, WANG Tao, WANG Xuping, et al. A Dual-threshold SVDD online fault diagnosis algorithm with faulty samples and its application[J]. Journal of Vibration Engineering, 2016, 29(6):555-560. http://d.old.wanfangdata.com.cn/Periodical/zdgcxb201603023 李凌均, 张周锁, 何正嘉.基于支持向量数据描述的机械故障诊断研究[J].西安交通大学学报, 2003, 37(9):910-913. doi: 10.3321/j.issn:0253-987X.2003.09.008LI Lingjun, ZHANG Zhousuo, HE Zhengjia. Research of mechanical system fault diagnosis based on support vector data description[J]. Journal of Xi'an Jiaotong University, 2003, 37(9):910-913. doi: 10.3321/j.issn:0253-987X.2003.09.008 张莹.数控机床机械故障诊断方法的研究及应用[D].大连: 大连理工大学, 2005. http://www.cnki.com.cn/Article/CJFDTOTAL-SDGJ201715017.htm TAX D M J, DUIN R P W. Support vector data description[J]. Machine Learning, 2004, 54(1):45-66. http://d.old.wanfangdata.com.cn/Periodical/xajtdxxb200309008 THIEULLEN A, OULADSINE M, PINATON J. A survey of health indicators and data-driven prognosis in semiconductor manufacturing process[J]. IFAC Proceedings Volumes, 2012, 45(20):19-24. doi: 10.3182/20120829-3-MX-2028.00246 潘志松, 陈斌, 缪志敏, 等. One-Class分类器研究[J].电子学报, 2009, 37(11):2496-2503. doi: 10.3321/j.issn:0372-2112.2009.11.025PAN Zhisong, CHEN Bin, LIAO Zhimin, et al. Overview of study on one-class classifiers[J]. Chinese Journal of Electronics, 2009, 37(11):2496-2503. doi: 10.3321/j.issn:0372-2112.2009.11.025 罗隽, 潘志松, 缪志敏, 等.基于写相关支持向量描述的入侵防护审计模型研究[J].通信学报, 2007, 28(7):8-14. doi: 10.3321/j.issn:1000-436x.2007.07.002LUO Jun, PAN Zhisong, LIAO Zhimin, et al. Research on the security audit model in intrusion prevention based on write-related support Vector data description[J]. Journal on Communications, 2007, 28(7):8-14. doi: 10.3321/j.issn:1000-436x.2007.07.002 ZATTONI E.Detection of incipient failures using an H2-norm criterion:application to electric point machines[J]. Ifac World Congress, 2005, 38(1):250-255. http://www.sciencedirect.com/science/article/pii/S1474667016377825 汪倩.面向PHM技术应用的转辙机数据采集系统构建与验证[D].广州: 暨南大学, 2014. http://cdmd.cnki.com.cn/Article/CDMD-10559-1015002886.htm ARDAKANI H D, LUCAS C, SIEGEL D, et al. PHM for railway system-A case study on the health assessment of the point machines[C]//2012 IEEE Conference on Prognostics and Health Management (PHM).[S.l.]: IEEE, 2012: 1-5. -
下载:
点击查看大图
图(9)
计量
- 文章访问数: 435
- HTML全文浏览量: 208
- PDF下载量: 74
- 被引次数: 0