小波-形态-EMD综合分析法及其应用  

HYBRID WAVELET-MORPHOLOGY-EMD ANALYSIS AND ITS APPLICATION

作　　者：柏林[1] 刘小峰[1] 秦树人[1] 

机构地区：[1]重庆大学机械学院测试中心,重庆400044

出　　处：《振动与冲击》2008年第5期1-4,35,共5页Journal of Vibration and Shock

基　　金：国家自然科学基金资助项目(No50605065);重庆市自然科学基金资助项目(No2007BB2142)

摘　　要：在经验模态分解(EMD)的理论基础上,分析了随机白噪声及局部强干扰对EMD分解质量的影响,结合小波消噪和形态滤波理论,提出了一种新型的小波-形态-EMD算法模型。该模型将小波形态变换作为EMD前置滤波单元,可以减少不必要的分解层次,降低EMD分解的边界积累效应,消除局部强干扰造成的模态裂解现象,有效提高EMD的时效性和精确度。利用仿真信号分析实例详述了这种综合分析方法的实施过程,并将该方法成功运用于齿轮故障的早期检测中。实验结果证明该方法在机械故障诊断中切实可行,具有较好的应用价值。The principle of empirical mode decomposition (EMD) is briefly introduced, and the influence caused by random white noises and local strong disturbances embedded in signal on EMD results is discussed. Aiming at increasing the precision and effectiveness of EMD, a novel integrated wavelet-morphology filter-EMD method is presented based on the principles of wavelet denoising and morphology filtering. In the method, the wavelet transform combining with mathematical morphology transform is taken as the pre-filter process unit in order to reduce the unnecessary decomposition levels and boundary accumulated errors and remove mode mixing phenomenon caused by local stronger disturbances. The process of this integrated method is described in detail using a simulated signal analysis example. Moreover the method is applied in the early stage fault diagnosis of gearbox and the results verify its practicality and validity.

关 键 词：经验模态分解 小波消噪 形态滤波 边界积累误差 模态混叠 

分 类 号：TG156[金属学及工艺—热处理]