形态学滤波和HHT变换在轮轨故障诊断中的应用  

Application of morphological filter and HHT transform in wheel rail fault diagnosis

作　　者：李大柱 Li Dazhu(China Railway First Surveying and Design Institute Group Co.,Ltd.,Shaanxi Xi'an,710043,China)

机构地区：[1]中铁第一勘察设计院集团有限公司,陕西西安710043

出　　处：《机械设计与制造工程》2025年第1期95-99,共5页Machine Design and Manufacturing Engineering

摘　　要：针对现有轮轨故障诊断装置结构复杂、成本高且实时性差等问题,提出了一种基于轴箱振动加速度响应的轮轨故障诊断方法。首先通过形态学滤波器对轴箱振动信号进行降噪处理,然后运用集合经验模态分解(EEMD)将处理后的信号分解得到若干阶固有模态函数(IMF),再依据能量熵增量的相对大小剔除IMF分量中的虚假分量,对剩余的有效分量进行希尔伯特-黄变换(HHT)得到Hilbert谱。研究结果表明,车辆运行在正常工况与不同类型轮轨故障工况下,轴箱振动加速度的Hilbert谱有显著的差异,因此依据Hilbert谱的特征可有效诊断轮轨故障。Aiming at complex structure,high cost and poor real-time performance of the existing wheel-rail fault diagnosis device,a wheel-rail fault diagnosis method based on transmission box vibration acceleration response is proposed.Firstly,the transmission box vibration signal is denoised by morphological filter,and then the processed signal is decomposed by ensemble empirical mode decomposition(EEMD)to obtain several intrinsic mode functions(IMF),and then the false components in the IMF components are eliminated according to the relative magnitude of the energy entropy increment,and the Hilbert-Huang transform(HHT)is used to obtain the Hilbert spectrum by the remaining effective components.The results show that there is a significant difference between the Hilbert spectrum of transmission box acceleration under normal conditions and different types of wheel-rail faults,so the Hilbert spectrum can be used to effectively diagnose wheel-rail faults.

关 键 词：轮轨故障 轴箱振动加速度 形态学滤波 集合经验模态分解 希尔伯特-黄变换 

分 类 号：U279.2[机械工程—车辆工程]