摘要
针对随机噪声和虚假分量影响总体平均经验模态分解(EEMD)分解质量问题,提出基于奇异值分解(SVD)和第二代小波变换(SGWT)联合降噪预处理和本征模态分量(IMF)能量熵增量剔除虚假分量的改进EEMD方法。该方法首先对原始信号进行第二代小波变换,利用SVD对SGWT得到的高频系数进行降噪处理,克服了软、硬阈值法降噪的缺陷。然后对消噪处理的信号进行EEMD分解,通过IMF能量熵增量去除虚假分量;最后对主IMF分量进行Hilbert谱分析来提取信号的主要特征。仿真和实验结果表明,SVD和SGWT联合降噪故障信号信噪比显著提高,且失真度小,抑制了噪声对EEMD分解精度的干扰,能量熵增量能有效地去除虚假IMF,Hilbert谱中各频率成分清晰不混叠,成功提取了液压系统故障特征频率。
For the problem that random noise and false intrinsic mode function （IMF） decline the quality of EEMD decomposition, an improved ensemble empirical mode decomposition （EEMD） method is presented based on singular value decomposition （SVD） and second generation wavelet transform （SGWT） to de-noising pre-processing and EEMD energy entropy increment to remove the false IMFs. Firstly, the original signal is processed by SGWT. SVD is applied to de-noise the high frequency coefficients, which overcomes the defect of soft and hard threshold method. Secondly, de-noised signal is decomposed through EEMD and IMF is used to remove the false component. Finally, the main IMFs are analyzed by the Hilbert spectrum. Simulation and experimental result show that the SVD-SGWT de-nosing can not only significantly increase signal to noise ratio and have less distortion but also depress the noise impact of the accuracy of EEMD. Energy entropy increment can effectively remove the false IMFs. The each frequency of Hilbert spectrum is clear and the method proposed effectively extracts the fanlts feature frequency of hydraulic syste~
出处
《机械设计与制造》
北大核心
2015年第3期51-54,共4页
Machinery Design & Manufacture
基金
2011年国家自然科学基金项目(51175511)
关键词
奇异值分解
第二代小波变换
总体平均经验模态分解
固有模态函数
能量熵增量
故障特征提取
Singular Value Decomposition
Second Generation Wavelet Transform
Ensemble Empirical Mode Deco-mposition
False Intrinsic Mode Function
Energy Entropy Increment
Fault Feature Extraction
