基于全矢谱的全信息能量研究
|
摘要
随着现代旋转机械设备向大型化、复杂化、自动化、高速化、重载化的方向发展,状态监测和故障诊断技术在国民经济基础领域中发挥着越来越重要的作用。转子的振动信息通常由截面内相互垂直安装的传感器采集,但传统的旋转机械状态监测和故障诊断只采用单源振动信号作为判别设备运行状态的依据,割裂了各个通道信号之间的有机联系,导致信息不完整、不可靠。由于单源信号从量值和结构方面都很难反映设备运行的实际状况,因此采用传统的分析方法很容易造成漏判和误判。全信息技术实现了旋转机械振动信息的无遗漏检测,极大改善了旋转机械状态监测和故障诊断建立在不完整信息基础之上的现状,实践证明具有巨大的优越性。
本文以多传感器信号的数据级融合和转子动力学基本理论为基础,介绍了基于信息融合的全矢谱技术理论基础、数值方法、图谱表达、物理意义及其在故障诊断工作中的应用。全矢谱分析技术具有全面、直观、易拓展的特点,反映了转子的真实运动特征。
在融合信号的矢量谱分析基础之上,结合传统的信号功率谱分析理论,探讨了多传感器融合信号的全矢功率谱定义、性质、实现途径以及其谱估计方法。研究表明,全矢功率谱具有清晰的物理意义,改善了传统单通道信号功率谱信息不完整、反映问题片面的缺陷,谱估计方法丰富多样,计算简捷方便,为故障诊断工作提供了可靠的依据。
为了提高全矢谱分析的分辨率,将复调制频谱细化原理与全矢谱理论相结合,多传感器融合信号经过复调制频移—低通滤波—选抽—快速傅里叶变换等步骤之后,实现了融合矢量信号频谱的局部细化。全矢细化谱兼具信息全面和高分辨率的特点,运算量远小于具有相同分辨率的普通全矢谱,能够综合反映融合矢量信号的频域局部细微特征,有利于提高故障诊断工作的效率及准确性,并节约硬件成本。该方法可应用于旋转机械故障诊断中,对频谱密集型多源融合矢量信号进行有效分析。
结合全信息技术、小波分析以及信息熵的基本理论,利用小波变换将两个垂直通道的信号分别分解至不同的频率带,综合所有分解系数计算得到全矢小波能量熵,对融合信号能量分布的紊乱程度进行量化。全矢小波能量熵能反映融合振动信号能量分布的复杂性,且对能量分布的变化较为敏感,能够检测故障引起的信号异常,并可利用其对故障的发展趋势进行预测,从而可作为衡量设备工作状态的指标应用到旋转机械状态监测领域。
在理论探讨的基础上,通过在matlab环境下进行编程验证,表明本文所提出的新方法有效、实用。同时也证明全矢谱分析能够真实反映机械振动的全部特征、极大提高了诊断的客观性和准确性,是一种具有极高理论研究价值和工程应用价值、发展前景广阔的技术。
With the maximization, complication, automation, high-speed and heavy-load orientation of the modern rotating machinery, the technologies in terms of condition monitoring and fault diagnosis are playing an increasingly important role in the fundamental areas of national industry.The information of the rotor is usually collected by the sensors vertically installed on the section. However, as for the traditional condition monitoring and fault diagnosis of rotating machinery only base on the single-source vibration signal and disserve the organic relations among each channel,thus it leads to the incomplete and unreliable information.It is diffcult for the single-source signal to reflect the work condition of the equipment,and as a result, it is easy to lose or get mistaken conclusions.The full information technologies have successfully made the diagnosis results come true. They have greatly improved the condition monitoring and fault diagnosis of rotation machinery. Practices show it is of great superiority.
On the basis of the fundamental theories about data fusion and dynamics of rotor, the basic theory of full vector spectrum, and its numerical method, spectrum illustration, physical meaning and application in fault diagnosis are well introduced. Full vector spectrum analysis technology is featured by being comprehensive, direct and easy to be exploited. It can reflect the real work condition of the rotor.
Based on the full vector spectrum theory of fusion signals together with traditional signal power spectrum analysis theory, the definition, quality and implement method of full vector power spectrum of multi-sensor fusion signal as well as its estimation method are discussed in details. Studies show that full vector power spectrum possesses explicit physical meaning, improves the defects of the traditional single-source signal power spectrum technology which used to show incomplete and one-sided information, thus providing the only evidence for fault diagnosis.
In order to improve the resolution of the full vector spectrum, it is to integrate the zoom spectrum theory of complex modulation and full vector spectrum theory.After several steps like complex modulation frequency-shifting,low-pass filtering,sub-sampling and FFT,the vector spectrum of the multi-sensor fusion signal has been partially zoomed.Full vector zoom spectrum is characterized by showing comprehensive information and high resolution. Its computation load is far less than the common full vector spectrum with the same resolution. It is able to fully reflect the partial fine features of frequency domain of fusion vector signal, and improve the efficiency and accuracy of fault diagnosis and effectively save the cost of the hardware.It can be applied in fault diagnosis of rotation machinery and effectively analyze multi-resource vector signal with intensive frequency.
With the basis theories in terms of full information technology, wavelet analysis and information entropy, the signals in two vertical channels can be decomposed into two different frequency bands by wavelet. Then full vector wavelet energy entropy can be reached by calculating all decomposition coefficients, thus quantizing the disorder degree of energy distribution of fusion signal. Full information wavelet energy entropy can fully show the complexity of energy distribution of the fusion vibration signal and be sensitive to the changes of energy distribution, thus successfully predicting the fault and its development trend. It can be used as a standard for judging equipment operation and be used in condition monitoring of rotation machinery.
After careful theory discussion and programming test under the matlab enviroment, the new method above is proved to be effective and practical. At the same time, it shows that the full vector spectrum analysis method can truly reflect the whole features of machinery vibration and greatly improve the objectivity and accuracy of diagnosis. It is of a new technology with high research value, engineering application and broad development prospect.
本文以多传感器信号的数据级融合和转子动力学基本理论为基础,介绍了基于信息融合的全矢谱技术理论基础、数值方法、图谱表达、物理意义及其在故障诊断工作中的应用。全矢谱分析技术具有全面、直观、易拓展的特点,反映了转子的真实运动特征。
在融合信号的矢量谱分析基础之上,结合传统的信号功率谱分析理论,探讨了多传感器融合信号的全矢功率谱定义、性质、实现途径以及其谱估计方法。研究表明,全矢功率谱具有清晰的物理意义,改善了传统单通道信号功率谱信息不完整、反映问题片面的缺陷,谱估计方法丰富多样,计算简捷方便,为故障诊断工作提供了可靠的依据。
为了提高全矢谱分析的分辨率,将复调制频谱细化原理与全矢谱理论相结合,多传感器融合信号经过复调制频移—低通滤波—选抽—快速傅里叶变换等步骤之后,实现了融合矢量信号频谱的局部细化。全矢细化谱兼具信息全面和高分辨率的特点,运算量远小于具有相同分辨率的普通全矢谱,能够综合反映融合矢量信号的频域局部细微特征,有利于提高故障诊断工作的效率及准确性,并节约硬件成本。该方法可应用于旋转机械故障诊断中,对频谱密集型多源融合矢量信号进行有效分析。
结合全信息技术、小波分析以及信息熵的基本理论,利用小波变换将两个垂直通道的信号分别分解至不同的频率带,综合所有分解系数计算得到全矢小波能量熵,对融合信号能量分布的紊乱程度进行量化。全矢小波能量熵能反映融合振动信号能量分布的复杂性,且对能量分布的变化较为敏感,能够检测故障引起的信号异常,并可利用其对故障的发展趋势进行预测,从而可作为衡量设备工作状态的指标应用到旋转机械状态监测领域。
在理论探讨的基础上,通过在matlab环境下进行编程验证,表明本文所提出的新方法有效、实用。同时也证明全矢谱分析能够真实反映机械振动的全部特征、极大提高了诊断的客观性和准确性,是一种具有极高理论研究价值和工程应用价值、发展前景广阔的技术。
With the maximization, complication, automation, high-speed and heavy-load orientation of the modern rotating machinery, the technologies in terms of condition monitoring and fault diagnosis are playing an increasingly important role in the fundamental areas of national industry.The information of the rotor is usually collected by the sensors vertically installed on the section. However, as for the traditional condition monitoring and fault diagnosis of rotating machinery only base on the single-source vibration signal and disserve the organic relations among each channel,thus it leads to the incomplete and unreliable information.It is diffcult for the single-source signal to reflect the work condition of the equipment,and as a result, it is easy to lose or get mistaken conclusions.The full information technologies have successfully made the diagnosis results come true. They have greatly improved the condition monitoring and fault diagnosis of rotation machinery. Practices show it is of great superiority.
On the basis of the fundamental theories about data fusion and dynamics of rotor, the basic theory of full vector spectrum, and its numerical method, spectrum illustration, physical meaning and application in fault diagnosis are well introduced. Full vector spectrum analysis technology is featured by being comprehensive, direct and easy to be exploited. It can reflect the real work condition of the rotor.
Based on the full vector spectrum theory of fusion signals together with traditional signal power spectrum analysis theory, the definition, quality and implement method of full vector power spectrum of multi-sensor fusion signal as well as its estimation method are discussed in details. Studies show that full vector power spectrum possesses explicit physical meaning, improves the defects of the traditional single-source signal power spectrum technology which used to show incomplete and one-sided information, thus providing the only evidence for fault diagnosis.
In order to improve the resolution of the full vector spectrum, it is to integrate the zoom spectrum theory of complex modulation and full vector spectrum theory.After several steps like complex modulation frequency-shifting,low-pass filtering,sub-sampling and FFT,the vector spectrum of the multi-sensor fusion signal has been partially zoomed.Full vector zoom spectrum is characterized by showing comprehensive information and high resolution. Its computation load is far less than the common full vector spectrum with the same resolution. It is able to fully reflect the partial fine features of frequency domain of fusion vector signal, and improve the efficiency and accuracy of fault diagnosis and effectively save the cost of the hardware.It can be applied in fault diagnosis of rotation machinery and effectively analyze multi-resource vector signal with intensive frequency.
With the basis theories in terms of full information technology, wavelet analysis and information entropy, the signals in two vertical channels can be decomposed into two different frequency bands by wavelet. Then full vector wavelet energy entropy can be reached by calculating all decomposition coefficients, thus quantizing the disorder degree of energy distribution of fusion signal. Full information wavelet energy entropy can fully show the complexity of energy distribution of the fusion vibration signal and be sensitive to the changes of energy distribution, thus successfully predicting the fault and its development trend. It can be used as a standard for judging equipment operation and be used in condition monitoring of rotation machinery.
After careful theory discussion and programming test under the matlab enviroment, the new method above is proved to be effective and practical. At the same time, it shows that the full vector spectrum analysis method can truly reflect the whole features of machinery vibration and greatly improve the objectivity and accuracy of diagnosis. It is of a new technology with high research value, engineering application and broad development prospect.
引文
[1]韩捷,张瑞林.旋转机械故障机理及诊断技术[M].北京:机械工业出版社,1997.
[2]陈大禧,朱铁光.大型回转机械诊断现场实用技术[M].北京:机械工业出版社,2002.
[3]徐章遂.故障信息诊断原理及应用[M].北京:国防工业出版社,2000.
[4]钟秉林,黄仁.机械故障诊断学[M].北京:机械工业出版社,2006.
[5]汪江萍.机械设备故障诊断技术及应用[M].西安:西北工业大学出版社,2001.
[6]陈长征,胡立新,周勃,等.设备振动分析与故障诊断技术[M].北京:科学出版社,2007.
[7]虞和济,韩庆大,李沈.设备故障诊断工程[M].北京:冶金工业出版社,2001.
[8]韩捷,关惠玲,梁川,等.矢谱:一种实用的旋转机械故障诊断分析方法[J].机械强度,1998,20(3):212-215.
[9]金敬强,武富春.信息融合技术的发展现状与展望[J].电脑开发与应用,2006,19(1):50-58.
[10]晋风华,李录平.多传感器信息融合技术及其在旋转机械振动故障诊断中的应用[J].热力发电,2004,(5):45-48.
[11]董志荣.论信息融合[J].情报指挥控制系统与仿真技术,2001,(7):27-36.
[12]刘同明,夏祖勋,解洪成.数据融合技术及其应用[M].北京:国防工业出版社,1998.
[13]潘泉,于昕,程咏梅,等.信息融合理论的基本方法与进展[J].自动化学报,2003,29(4):599-615.
[14]徐毅,金德琨,敬忠良.数据融合研究的回顾与展望[J].信息与控制,2002,31(3):250-253.
[15]Blackman.Multiple-Target Tracking with Radar Applications[M].Dedham:Artch house,INC.,Norwood,MA,1986.
[16]Torfinn Taxt.,Anne H.Schistad Solberg.Information fusion in remote sensing.Vistas in Astronomy,1997,41(3):337-342.
[17]王耀南,李树涛.多传感器信息融合及应用综述[J].控制与决策,2001,16(5):518-522.
[18]王丽雅.基于信息融合的旋转机械全信息时频分析方法研究[D].郑州大学硕士学位论文,2005.
[19]程军圣,于德介,杨宁,等.尺度-小波能量谱在滚动轴承故障诊断中的应用[J].振动工程学报,2004,17(1):82-85.
[20]张建伟,相敬林.使用小波理论分析信号功率谱的二次处理方法[J].信号处理,1999,15(2):125-128.
[21]李辉,郑海起,唐力伟.基于EMD和功率谱的齿轮故障诊断研究[J].振动与冲击,2006,25(1):133-135.
[22]Qu Liangsheng,Liu Xiong,Chen Yaodong.The Holospectrum:A New Method for Rotor Surveillance and Diagnosis[J].Mechanical Signature&Signal Processing,1989,8(8),25B-288.
[23]屈梁生,史东锋.全息谱十年:回顾与展望[J].振动、测试与诊断,1998,18(4):235-242.
[24]刘雄,屈梁生.基于二维全息谱的回转机械亚同步振动分析[J].化工机械,1994,21(1):34-37.
[25]史东锋,屈梁生.三维全息谱分解在回转机械诊断中的应用研究[J].西安交通大学学报,1998,32(7):94-97.
[26]Goldman P.,Muszynska A..Application of full spectrum to rotating machinery diagnostics[P].Orbit,Vol.,No.1,First Quarter 1999,Bently Nevada Corporation.
[27]韩捷,石来德.全谱分析技术及其在故障诊断中的应用[J].郑州大学学报,2002,23(3):44-47.
[28]钟一谔,何衍宗,王正,等.转子动力学[M].北京:清华大学出版社,1987.
[29]韩捷,董辛旻,郝伟,等.旋转机械同源数据全信息倒频谱分析与应用[J].机械强度,2005,27(4):452-455.
[30]韩捷,石来德.旋转机械的全信息能量谱分析方法研究[J].机械强度,2003,25(4):364-368.
[31]胡广书.数字信号处理理论算法与实现(第二版)[M].北京:清华大学出版社,2003.
[32]张贤达.现代信号处理(第二版)[M].北京:清华大学出版社,2002.
[33]毕果,韩捷,梁川.基于矢量振动信号的AR功率谱分析及应用[J].郑州大学学报,2003,24(2):80-83.
[34]飞思科技产品研发中心.MATLAB7辅助信号处理技术及应用[M].北京:电子工业出版社,2005.
[35]郑方,徐明星.信号处理原理[M].北京:清华大学出版社,2003.
[36]孔德仁,朱蕴璞,狄长安.工程测试与信息处理[M].北京:国防工业出版社,2003:350-352.
[37]丁康,谢明,张彼德,等.基于复解析带通滤波器的复调制细化谱分析原理和方法[J].振动工程学报,2001,14(1):29-35.
[38]Runtz K J,Hack D.A multistage DFT-FFT-CZT approach for accurate efficient analysis of sparsely distributed spectra[C].//Ken Ferens.Proceedings of the 2002 IEEE Canadian Conference on Electrical & Computer Engineering.IEEE Canadian Conference on Electrical and Computing Engineering.Winnipeg,Manitoba,Canada.2002,Piscataway,NJ,USA:IEEE Service Centre,2002:127-132 vol.1.
[39]Hoyer E A,Stork R F.The zoom FFT using complex modulation[C].//IEEE.IEEE Proceedings of ICASP'77.IEEE International Conference on Acoustics,Speech and Signal Processing Hartford,Connecticut.1977,Piscataway,NJ,USA:IEEE Press,1977:78-81 vol.2.
[40]C W Nawrocki.ZOOM-FFT-An Approximate Vernier Frequncy Algorithm[R].TM-SA2201-584-72,NUSC,22 Nov,1972.
[41]R C Y Yip.Some aspects of the zoom transform[J].IEEE Trans on Computers,1976,25(3):287-296.
[42]A H Nuttall.An Approximate Fast Fourier Transform Technique for Vernier Spectral Analysis[R].NUSC,TR4767,1974.
[43]丁康,李巍华,朱小勇.齿轮及齿轮箱故障诊断实用技术[M].北京:机械工业出版社,2005.
[44]罗利春.Zoom-FFT的改进、频谱反演与时-频局部化特性[J].电子学报,2006,34(1):77-82.
[45]石峰,莫忠息.信息论基础(第二版)[M].武昌:武汉大学出版社:2006.
[46]李弼成,罗建书.小波分析及其应用[M].北京:电子工业出版社,2003:1-86.
[47]Rosso O A,Blanco S,Yordanova J,et al.Wavelet entropy:a new tool for analysis of short duration brain electrical signals[J].Journal of Neuroscience Methods,2001,105(1):65-75.
[48]Daubechies I.Ten Lecture on Wavelet[M].Philadelphia:Society for Industrial and Applied Mathematics,1992.
[49]彭玉华.小波变换与工程应用[M].北京:科学出版社,1999.
[50]Mallat S,Hwang W L.Singularity detection and processing with wavelet[J].IEEE Transactions on Information Theory,1992,38(2):617-643.
[51]Nowak R D,Baraniuk R G.Wavelet-based transformations for nonlinear signal processing.IEEE Transactions on Signal Processing,1999,47(7):1852-1865.
[52]Daubechies I.The wavelet transform,time-frequency-localization and signal analysis[J].IEEE Trans IT,1990.36(5):605-612.
[53]Daubechies I.Where do wavelet come from?A personal point of view[J].Proc of IEEE,1996,84(4):510-513.
[54]杨建国.小波分析及其工程应用[M].北京:机械工业出版社,2005.
[55]何正友,陈小勤.基于多尺度能量统计和小波能量熵测度的电力暂态信号识别方法[J].中国电机 工程学报,2006,26(10):33-39.
[56]封洲燕.应用小波熵分析大鼠脑电信号的动态变化特性[J].生物物理学报,2002,18(3):325-330.
[57]肖余粮,和卫星,陈小平,等小波变换和小波熵在睡眠脑电信号变化特性研究中的应用价值[J].中国临床康复,2006,10(25):118-120.
[58]冯彩虹.全信息小波分析及设备诊断工程应用研究[D].郑州大学硕士学位论文,2006.
[59]朱雪龙.应用信息论基础[M].北京:清华大学出版社,2001.
[60]李建勋,柯熙政,丁德强.一种基于小波熵的时间尺度算法[J].天文学报,2007,48(1):84-91.
[61]中弢,黄树红,韩守木,等.旋转机械振动信号的信息熵特征[J].机械工程学报,2001,37(6):94-98.
[62]颜志国,王志中,任小梅.独立分量分析和小波熵在动作模式分类中的应用[J].北京生物医学工程,2006,25(5):457-465.
[63]印欣运,何永勇,彭志科,等.小波熵及其在状态趋势分析中的应用[J].振动工程学报,2004,17(2):165-169.
[2]陈大禧,朱铁光.大型回转机械诊断现场实用技术[M].北京:机械工业出版社,2002.
[3]徐章遂.故障信息诊断原理及应用[M].北京:国防工业出版社,2000.
[4]钟秉林,黄仁.机械故障诊断学[M].北京:机械工业出版社,2006.
[5]汪江萍.机械设备故障诊断技术及应用[M].西安:西北工业大学出版社,2001.
[6]陈长征,胡立新,周勃,等.设备振动分析与故障诊断技术[M].北京:科学出版社,2007.
[7]虞和济,韩庆大,李沈.设备故障诊断工程[M].北京:冶金工业出版社,2001.
[8]韩捷,关惠玲,梁川,等.矢谱:一种实用的旋转机械故障诊断分析方法[J].机械强度,1998,20(3):212-215.
[9]金敬强,武富春.信息融合技术的发展现状与展望[J].电脑开发与应用,2006,19(1):50-58.
[10]晋风华,李录平.多传感器信息融合技术及其在旋转机械振动故障诊断中的应用[J].热力发电,2004,(5):45-48.
[11]董志荣.论信息融合[J].情报指挥控制系统与仿真技术,2001,(7):27-36.
[12]刘同明,夏祖勋,解洪成.数据融合技术及其应用[M].北京:国防工业出版社,1998.
[13]潘泉,于昕,程咏梅,等.信息融合理论的基本方法与进展[J].自动化学报,2003,29(4):599-615.
[14]徐毅,金德琨,敬忠良.数据融合研究的回顾与展望[J].信息与控制,2002,31(3):250-253.
[15]Blackman.Multiple-Target Tracking with Radar Applications[M].Dedham:Artch house,INC.,Norwood,MA,1986.
[16]Torfinn Taxt.,Anne H.Schistad Solberg.Information fusion in remote sensing.Vistas in Astronomy,1997,41(3):337-342.
[17]王耀南,李树涛.多传感器信息融合及应用综述[J].控制与决策,2001,16(5):518-522.
[18]王丽雅.基于信息融合的旋转机械全信息时频分析方法研究[D].郑州大学硕士学位论文,2005.
[19]程军圣,于德介,杨宁,等.尺度-小波能量谱在滚动轴承故障诊断中的应用[J].振动工程学报,2004,17(1):82-85.
[20]张建伟,相敬林.使用小波理论分析信号功率谱的二次处理方法[J].信号处理,1999,15(2):125-128.
[21]李辉,郑海起,唐力伟.基于EMD和功率谱的齿轮故障诊断研究[J].振动与冲击,2006,25(1):133-135.
[22]Qu Liangsheng,Liu Xiong,Chen Yaodong.The Holospectrum:A New Method for Rotor Surveillance and Diagnosis[J].Mechanical Signature&Signal Processing,1989,8(8),25B-288.
[23]屈梁生,史东锋.全息谱十年:回顾与展望[J].振动、测试与诊断,1998,18(4):235-242.
[24]刘雄,屈梁生.基于二维全息谱的回转机械亚同步振动分析[J].化工机械,1994,21(1):34-37.
[25]史东锋,屈梁生.三维全息谱分解在回转机械诊断中的应用研究[J].西安交通大学学报,1998,32(7):94-97.
[26]Goldman P.,Muszynska A..Application of full spectrum to rotating machinery diagnostics[P].Orbit,Vol.,No.1,First Quarter 1999,Bently Nevada Corporation.
[27]韩捷,石来德.全谱分析技术及其在故障诊断中的应用[J].郑州大学学报,2002,23(3):44-47.
[28]钟一谔,何衍宗,王正,等.转子动力学[M].北京:清华大学出版社,1987.
[29]韩捷,董辛旻,郝伟,等.旋转机械同源数据全信息倒频谱分析与应用[J].机械强度,2005,27(4):452-455.
[30]韩捷,石来德.旋转机械的全信息能量谱分析方法研究[J].机械强度,2003,25(4):364-368.
[31]胡广书.数字信号处理理论算法与实现(第二版)[M].北京:清华大学出版社,2003.
[32]张贤达.现代信号处理(第二版)[M].北京:清华大学出版社,2002.
[33]毕果,韩捷,梁川.基于矢量振动信号的AR功率谱分析及应用[J].郑州大学学报,2003,24(2):80-83.
[34]飞思科技产品研发中心.MATLAB7辅助信号处理技术及应用[M].北京:电子工业出版社,2005.
[35]郑方,徐明星.信号处理原理[M].北京:清华大学出版社,2003.
[36]孔德仁,朱蕴璞,狄长安.工程测试与信息处理[M].北京:国防工业出版社,2003:350-352.
[37]丁康,谢明,张彼德,等.基于复解析带通滤波器的复调制细化谱分析原理和方法[J].振动工程学报,2001,14(1):29-35.
[38]Runtz K J,Hack D.A multistage DFT-FFT-CZT approach for accurate efficient analysis of sparsely distributed spectra[C].//Ken Ferens.Proceedings of the 2002 IEEE Canadian Conference on Electrical & Computer Engineering.IEEE Canadian Conference on Electrical and Computing Engineering.Winnipeg,Manitoba,Canada.2002,Piscataway,NJ,USA:IEEE Service Centre,2002:127-132 vol.1.
[39]Hoyer E A,Stork R F.The zoom FFT using complex modulation[C].//IEEE.IEEE Proceedings of ICASP'77.IEEE International Conference on Acoustics,Speech and Signal Processing Hartford,Connecticut.1977,Piscataway,NJ,USA:IEEE Press,1977:78-81 vol.2.
[40]C W Nawrocki.ZOOM-FFT-An Approximate Vernier Frequncy Algorithm[R].TM-SA2201-584-72,NUSC,22 Nov,1972.
[41]R C Y Yip.Some aspects of the zoom transform[J].IEEE Trans on Computers,1976,25(3):287-296.
[42]A H Nuttall.An Approximate Fast Fourier Transform Technique for Vernier Spectral Analysis[R].NUSC,TR4767,1974.
[43]丁康,李巍华,朱小勇.齿轮及齿轮箱故障诊断实用技术[M].北京:机械工业出版社,2005.
[44]罗利春.Zoom-FFT的改进、频谱反演与时-频局部化特性[J].电子学报,2006,34(1):77-82.
[45]石峰,莫忠息.信息论基础(第二版)[M].武昌:武汉大学出版社:2006.
[46]李弼成,罗建书.小波分析及其应用[M].北京:电子工业出版社,2003:1-86.
[47]Rosso O A,Blanco S,Yordanova J,et al.Wavelet entropy:a new tool for analysis of short duration brain electrical signals[J].Journal of Neuroscience Methods,2001,105(1):65-75.
[48]Daubechies I.Ten Lecture on Wavelet[M].Philadelphia:Society for Industrial and Applied Mathematics,1992.
[49]彭玉华.小波变换与工程应用[M].北京:科学出版社,1999.
[50]Mallat S,Hwang W L.Singularity detection and processing with wavelet[J].IEEE Transactions on Information Theory,1992,38(2):617-643.
[51]Nowak R D,Baraniuk R G.Wavelet-based transformations for nonlinear signal processing.IEEE Transactions on Signal Processing,1999,47(7):1852-1865.
[52]Daubechies I.The wavelet transform,time-frequency-localization and signal analysis[J].IEEE Trans IT,1990.36(5):605-612.
[53]Daubechies I.Where do wavelet come from?A personal point of view[J].Proc of IEEE,1996,84(4):510-513.
[54]杨建国.小波分析及其工程应用[M].北京:机械工业出版社,2005.
[55]何正友,陈小勤.基于多尺度能量统计和小波能量熵测度的电力暂态信号识别方法[J].中国电机 工程学报,2006,26(10):33-39.
[56]封洲燕.应用小波熵分析大鼠脑电信号的动态变化特性[J].生物物理学报,2002,18(3):325-330.
[57]肖余粮,和卫星,陈小平,等小波变换和小波熵在睡眠脑电信号变化特性研究中的应用价值[J].中国临床康复,2006,10(25):118-120.
[58]冯彩虹.全信息小波分析及设备诊断工程应用研究[D].郑州大学硕士学位论文,2006.
[59]朱雪龙.应用信息论基础[M].北京:清华大学出版社,2001.
[60]李建勋,柯熙政,丁德强.一种基于小波熵的时间尺度算法[J].天文学报,2007,48(1):84-91.
[61]中弢,黄树红,韩守木,等.旋转机械振动信号的信息熵特征[J].机械工程学报,2001,37(6):94-98.
[62]颜志国,王志中,任小梅.独立分量分析和小波熵在动作模式分类中的应用[J].北京生物医学工程,2006,25(5):457-465.
[63]印欣运,何永勇,彭志科,等.小波熵及其在状态趋势分析中的应用[J].振动工程学报,2004,17(2):165-169.