为了实现经验小波变换中 Fourier 谱的自适应分割，提出了自适应无参经验小波变换(APEWT)方法。同时，为了克服希尔伯特变换解调的不足，更精确地估计信号的时频分布，提出了改进归一化希尔伯特变换(INHT)。通过分析仿真信号将 APEWT 和 INHT 方法与经验模态分解(EMD)、总体平均经验模态分解(EEMD)和局部特征尺度分解等方法进行对比，结果表明了 APEWT和INHT方法的优越性。最后，将基于APEWT和INHT的时频分析方法应用于转子局部碰磨故障诊断，试验数据分析结果表明，所提出的方法不仅能够有效地诊断转子局部碰磨故障，而且诊断效果优于E MD和E E MD方法。%To fulfill an adaptive separation of Fourier spectrum in EWT,an adaptive parameter-less EWT(APEWT)method was proposed herein.To overcome the shortcomings of Hilbert trans-form and estimate more accurate time-frequency distribution of a given signal,an improved normal-ized Hilbert transform(INHT)was put forward.The proposed APEWT and INHT were compared with empirical mode decomposition(EMD),ensemble EMD(EEMD)and local characteristic-scale de-composition(LCD)methods and the analysis results demonstrate the effectiveness of the proposed method.Finally,APEWT and INHT based time-frequency analysis method were applied to local rub-bing fault diagnosis of a rotor system,and the analysis results of experimental data indicate that the proposed method may fulfill rotor rubbing fault diagnosis effectively and have better effectiveness than that of EMD and EEMD methods.
展开▼
