集成全息希尔伯特谱分析及其在滚动轴承故障诊断中的应用  

Ensemble holo-Hilbert spectral analysis and its application in fault diagnosis of rolling bearing

作　　者：彭国良 郑近德[1] 潘海洋[1] 童靳于[1] 刘庆运[1] PENG Guoliang;ZHENG Jinde;PAN Haiyang;TONG Jinyu;LIU Qingyun(College of Mechanical Engineering,Anhui University of Technology,Ma’anshan 243032,China)

机构地区：[1]安徽工业大学机械工程学院,安徽马鞍山243032

出　　处：《振动与冲击》2024年第13期98-105,125,共9页Journal of Vibration and Shock

基　　金：国家自然科学基金项目(51975004);安徽省教育厅杰出青年基金项目(2022AH020032)。

摘　　要：全息希尔伯特谱分析(holo-Hilbert spectral analysis, HHSA)是一种新的信号解调分析处理技术,其采用双层经验模态分解(empirical mode decomposition, EMD),能够有效揭示非线性或非平稳振动信号中的跨尺度耦合关系。但是,EMD在信号分解过程中存在严重的模态混叠问题,导致瞬时频率估计不准确,影响了HHSA的分析精度。基于此,提出了集成全息希尔伯特谱分析(ensemble holo-Hilbert spectral analysis, EHHSA)方法。同时为了更精确解调故障特征信息,通过对载波变量进行积分,定义了一种可以揭示调制特征的调幅边际谱分析方法。最后,通过对滚动轴承仿真和实测数据进行分析,结果表明:与传统谱分析方法相比,所提EHHSA方法及调幅边际谱的特征提取性能和噪声鲁棒性更强。Holo-Hilbert spectral analysis(HHSA)is a new signal demodulation analysis and processing technique using double-layer empirical mode decomposition(EMD)to effectively reveal cross-scale coupling relations in nonlinear or non-stationary vibration signals.However,EMD suffers from severe mode aliasing during signal decomposition to cause inaccurate instantaneous frequency estimation,and affect the analysis accuracy of HHSA.Here,Ensemble Holo-Hilbert spectral analysis(EHHSA)method was proposed.At the same time,in order to more accurately demodulate fault feature information,an amplitude modulation marginal spectral analysis method being able to reveal modulation characteristics was defined by integrating carrier variables.Finally,the proposed method was used to analyze simulation data and measured data of rolling bearing.The results showed that compared with traditional spectral analysis methods,the proposed EHHSA method and amplitude modulation marginal spectrum have stronger feature extraction performance and noise robustness.

关 键 词：集成全息希尔伯特谱分析(EHHSA) 时频分析 集成经验模态分解(EEMD) 故障诊断 

分 类 号：TH165[机械工程—机械制造及自动化]