诊断铁路接触网硬点CEEMDAN-SPWVD边际指数方法  

Diagnosis Method for Hard Spot of Railway Catenary Based on CEEMDAN-SPWVD Marginal Index

作　　者：丁宇鸣 刘金朝 徐晓迪 张文轩 杨志鹏 王婧 DING Yuming;LIU Jinzhao;XU Xiaodi;ZHANG Wenxuan;YANG Zhipeng;WANG Jing(Infrastructure Inspection Research Institute,China Academy of Railway Sciences Corporation Limited,Beijing 100081,China)

机构地区：[1]中国铁道科学研究院集团有限公司基础设施检测研究所,北京100081

出　　处：《铁道学报》2022年第5期34-41,共8页Journal of the China Railway Society

基　　金：中国铁路总公司科技研究开发计划(J2018G006,P2018G051)。

摘　　要：在接触网检测中,通常用受电弓滑板滑行时垂直方向上的加速度最大值来表示硬点,基于这个特点,提出一种新的诊断铁路接触网硬点CEEMDAN-SPWVD边际指数方法。对受电弓垂向加速度数据进行自适应噪声的完整集合经验模态分解CEEMDAN,得到若干固有模态函数(IMF),计算各阶IMF的平滑伪Wigner Ville分布SPWVD。将各IMF的时频分布进行叠加得到弓网系统振动的时频信息。根据时频分布得到的边际谱移动窗频率选择范围,以及接触网硬点的冲击特性,选择合适的时间窗长,以计算信号的移动滤波边际谱,并用于诊断接触网硬点病害。根据大量的检测数据对移动边际谱进行归一化处理,得到接触网边际指数(CMI)。在该方法中CEEMDAN将同一时间不同成分信号分离开,以避免交叉项干扰,而SPWVD相比WVD分布能有效消除不同时间冲击之间产生的交叉项干扰,最后归一化处理得到的CMI解决了直接通过幅值诊断硬点绝对阈值难以确定的问题。将该方法应用于接触网检测车实测数据,经过一系列分析,验证了该方法的有效性和稳定性。In catenary detection,the maximum acceleration in vertical direction is usually used to represent the Hard spot.Based on this feature,a new Diagnosis Method for Hard spot of Railway Catenary Based on CEEMDAN-SPWVD Marginal Index is proposed.Firstly,complete EEMD with adaptive noise(CEEMDAN)is applied to the vertical acceleration data of pantograph to obtain several intrinsic mode functions(IMF),and the smoothed pseudo Wigner Ville distribution(SPWVD)of each order of IMF is calculated.The time-frequency information of pantograph catenary system is obtained by superimposing the time-frequency distribution of each IMF.Secondly,according to the marginal spectrum obtained from time-frequency distribution,the appropriate time window length is selected according to the impact characteristics of catenary hard spot,so as to calculate the mobile filter marginal spectrum of signals,and to diagnose Railway Catenary hard spot diseases.Finally,according to a large number of detection data,the moving marginal spectrum is normalized to obtain the catenary marginal index(CMI).In this method,CEEMDAN separates different components of signals at the same time to avoid cross term interference.Compared with WVD distribution,SPWVD can effectively eliminate the cross term interference between different time shocks.Finally,the CMI obtained by normalized processing solves the problem that it is difficult to determine the absolute threshold value directly through amplitude diagnosis.After a series of analysis,the validity and stability of the method are verified.

关 键 词：接触网 硬点诊断 时频分析 移动边际谱 受电弓垂向振动加速度 

分 类 号：U226.8[交通运输工程—道路与铁道工程]