Pack-Kurtogram

所属分类:matlab编程
开发工具:matlab
文件大小:219KB
下载次数:428
上传日期:2011-07-10 15:03:27
上 传 者jiangruihong
说明:  快速计算一维信号的塔式谱峭度图,选择谱峭度最大值区间,方便确定所需滤波器中心频率及带宽。
(Computes the fast kurtogram of signal x.This fast algorithm uses a pyramidal decomposition of the signal into a user-specified number of levels (a good choice is nlevel = 8). The first level is the classical kurtosis of the signal, the second level is the spectral kurtosis in 2 octaves, the third level the spectral kurtosis in 4 half-octaves, etc. Non-integer levels refer to divisions into third octaves. Here below is an illustration on how the frequency band is successively split when increasing the level)

文件列表:
Pack Kurtogram\Pack Kurtogram V3\binary.m (383, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\DBFB.m (793, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\demo_Fast_Kurtogram.m (479, 2011-05-19)
Pack Kurtogram\Pack Kurtogram V3\Fast_Kurtogram.m (6035, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\Find_stft_kurt.m (2671, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\Find_wav_kurt.m (2159, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\Kf_fft.m (999, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\Kf_W.m (2014, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\kurt.m (696, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\K_wpQ.m (1354, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\K_wpQ_filt.m (1480, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\K_wpQ_filt_local.m (925, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\K_wpQ_local.m (2430, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\max_IJ.m (481, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\raylinv.m (894, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\TBFB.m (535, 2011-07-10)
Pack Kurtogram\Pack Kurtogram V3\VOIE1.MAT (874144, 1999-09-10)
Pack Kurtogram\Pack Kurtogram V3 (0, 2011-07-10)
Pack Kurtogram (0, 2011-07-10)

{\rtf1\ansi\ansicpg1252\deff0\deflang1036{\fonttbl{\f0\froman\fprq2\fcharset0 Times New Roman;}{\f1\fswiss\fcharset0 Arial;}} {\*\generator Msftedit 5.41.15.1507;}\viewkind4\uc1\pard\f0\fs24 FAST_KURTOGRAM\tab\tab Computes the fast kurtogram of signal x\par \par \lang2057\par \ul\b Principle\ulnone\b0\par This fast algorithm uses a pyramidal decomposition of the signal into a user-specified number of levels (a good choice is nlevel = 8). The first level is the classical kurtosis of the signal, the second level is the spectral kurtosis in 2 octaves, the third level the spectral kurtosis in 4 half-octaves, etc. Non-integer levels refer to divisions into third octaves. Here below is an illustration on how the frequency band is successively split when increasing the level:\par \par |\'85\'85\'85\'85\'85\'85\'85\'85.\'85.\'85\'85..\'85\'85\'85\'85.|\tab\tab level 0\par |\'85\'85.\'85\'850\'85\'85.\'85\'85|\'85..\'85\'851\'85.\'85\'85|\tab\tab level 1\par |\'85\'85\'85..\'85\'85|\'85\'85.\'85\'85\'85|\'85\'85\'85.\'85\'85|\tab\tab level 1.6\par |\'85..00..\'85|\'85\'8501.\'85.|\'85.10\'85..|\'85..11..\'85|\tab\tab level 2\par |\'85....\'85|\'85.\'85.|\'85..\'85.|\'85...\'85|.\'85\'85.|\'85...\'85|\tab\tab level 2.6\par |.000.|.001.|.011.|.011.|.100.|.101.|.110.|.111.|\tab\tab level 3\par \par \tab\tab \'85 etc \'85\tab\par \tab\tab\tab\tab\tab\par \tab ---> frequency --->\par \par \par \ul\b Input arguments\ulnone\b0\par Type "help Fast_Kurtogram" at the Matlab prompt.\par \par \ul\b Outputs\ulnone\b0\par The kurtogram shows the kurtosis at each level and at each frequency on a color image. A high value of the kurtogram indicates a frequency band where the signal is impulsive. For example if the frequency band \ldblquote 10\rdblquote is selected, this means that filtering the signal with a band-pass filter of bandwidth Fe/4 and central frequency 0.3125Fe will enhance all impulses hidden in the signal. As such, it may be used to select the optimal pair of values (bandwidth \endash central frequency) of the band-pass filter that maximises the kurtosis of the filtered signal. \par \par \ul\b Usage\ulnone\b0\par Run the file "demo_kurtogram":\par \par \tab\b >> Do you want to prewhiten the signal ? (no = 0 ; yes = 1): 0\par \b0\par (Prewhitening the signal before computing the kurtogram often improves the results.)\par \par \tab\b >> Choose the kurtosis measure (classic = 1 ; robust = 2): 1\par \b0\par (Option 2 is to make the kurtogram more robust in certain instances where the signal is polluted by impulsive noise.)\par \par \tab\b >> Choose the algorithm (filterbank = 1 ; stft-based = 2): 1\par \b0\par (Two different fast algorithms.)\par \par \tab\b >> Do you want to filter out transient signals from the kurtogram (yes = 1 \tab ; no = 0): 1\par \b0\par (The kurtogram shows its highest value at level 6 and central frequency 0.0195 ; answer "1" if you want to bandpass filter the signal in that frequency band or in any other.)\par \par \tab\b >> Enter the optimal carrier frequency (btw 0 and 0.5) where to filter the \tab signal: .02\par \b0\par (Enter any frequency value -- as displayed on the horizontal axis -- that is included in the selected box of the kurtogram ; use the mouse to zoom on the image.)\par \par \tab\b >> Enter the optimal level (btw 0 and 7) where to filter the signal: 6\par \b0\par (Enter the corresponding level of decomposition as displayed on the vertical axis.)\par \par \tab\b >>Do you want to see the envelope spectrum (yes = 1 ; no = 0): 1\par \b0\par (Enter "1" if you want to see both the filtered signal and its envelope spectrum.)\par \par \tab\b >>Do you want to keep on filtering out transients (yes = 1 ; no = 0): 0\par \b0\par \par \ul\b References\par \ulnone\b0 J. Antoni, \'ab\~Fast Computation of the Kurtogram for the Detection of Transient Faults\~\'bb, Mechanical Systems and Signal Processing, Volume 21, Issue 1, January 2007, Pages 108-124.\par \par \par \par \lang1036\f1\fs20\par }

近期下载者

相关文件


收藏者