改进希尔伯特-黄变换方法在钢轨波磨检测中的应用  

The Application of Improved Hilbert-Huang Transform Method in Rail Wave Wear Detection

作　　者：吴传奇 柴晓冬[1] 李立明[1] 郑树彬[1] WU Chuanqi;CHAI Xiaodong;LI Liming;ZHENG Shubin(College of Urban Rail Transit,Shanghai University of Engineering Science,Shanghai 201620,China)

机构地区：[1]上海工程技术大学城市轨道交通学院,上海201620

出　　处：《铁道标准设计》2021年第9期74-80,共7页Railway Standard Design

基　　金：国家自然科学基金项目(51975347,51907117)。

摘　　要：针对现有钢轨波磨检测方式检测效率低,检测精度不高而致使检测难的情况,提出利用Gocator传感器获得钢轨波磨检测数据,再通过小波变换和改进希尔伯特-黄变换对检测数据进行处理及分析的新方法。该方法以钢轨波磨检测数据为研究对象,首先,采用Gocator传感器获得钢轨波磨检测数据;然后,使用含噪信号与降噪误差比法,将因车体振动等因素而产生的影响诊断准确性的噪声从所采数据中分离,利用小波变换和希尔伯特-黄变换对去除噪声后的数据进行处理及分析;最后结合时域和频域特征进行缺陷识别。通过仿真研究证明,含噪信号与降噪误差比法与总体平均经验模态分解的优越性。同时,实验研究表明,希尔伯特-黄变换相较于小波变换在钢轨波磨细微信号识别方面具有明显的优势,为钢轨波磨的诊断提供了一种切实可行的方法。In view of the low detection efficiency and accuracy of the existing rail wave wear detection methods,a new method is proposed to obtain the rail wave wear detection data with the Gocator sensor,and then to process and analyze the detection data by using wavelet transform and improved Hilbert-Huang transform.The method takes rail wave wear detection data as the research object.Firstly,the Gocator sensor is used to obtain the rail wave wear detection data.Then,the noise caused by car body vibration and other factors,which affects the diagnostic accuracy,is separated from the collected data by means of the error ratio method of noisy signal against noise reduction.The data after noise removal is processed and analyzed by way of Hilbert-Huang transform.Finally,defects are identified by combining time domain and frequency domain features.The advantages of the error ratio method and the overall average empirical mode decomposition are proved by simulation research.Meanwhile,research results show that the Hilbert-Huang transform has obvious advantages over wavelet transform in identifying micro signals of rail wave wear and provides a practical method for the diagnosis of rail wave wear.

关 键 词：钢轨波磨 Gocator传感器 小波变换 希尔伯特-黄变换 检测 

分 类 号：U216.3[交通运输工程—道路与铁道工程]