基于多传感器时频分布的机械故障信号欠定盲源分离方法
|
摘要
结合多传感器时频分布(multisensor time-frequency distributions,MTFD)和盲源分离(blind source separation,BSS)的特点,提出一种针对机械复合故障信号的欠定盲源分离方法。首先利用Wigner-Ville分布将观测信号转化为MTFD矩阵;然后对该矩阵进行白化处理和噪声阈值处理,并对其自动项进行选择,对其特征向量进行集群处理,从而得到源信号TFD的估计;最后对源信号进行重建,得到源信号的估计。仿真及试验结果表明,本文所提出的方法在处理非平稳复合信号的欠定盲源分离方面具有很好的效果。
Taking advantages of multisensor time-frequency distribution(MTFD)and blind source separation(BSS),this paper presents an underdetermined blind source separation method for composite mechanical fault signals.Firstly,the observed signal is converted into an MTFD matrix by using Wigner-Ville distribution.Secondly,whitening and noise threshold processes are performed on the newly constructed matrix,and then the matrix's automatic items are selected and its eigenvectors are clustered in order to estimate the time-frequency distribution of source signal.Finally,the estimated source signal is obtained by reconstruction.Simulation and experimental results show that the proposed method is fairly effective on underdetermined blind source separation of non-stationary composite signals.
Taking advantages of multisensor time-frequency distribution(MTFD)and blind source separation(BSS),this paper presents an underdetermined blind source separation method for composite mechanical fault signals.Firstly,the observed signal is converted into an MTFD matrix by using Wigner-Ville distribution.Secondly,whitening and noise threshold processes are performed on the newly constructed matrix,and then the matrix's automatic items are selected and its eigenvectors are clustered in order to estimate the time-frequency distribution of source signal.Finally,the estimated source signal is obtained by reconstruction.Simulation and experimental results show that the proposed method is fairly effective on underdetermined blind source separation of non-stationary composite signals.
引文
[1]焦卫东,杨世锡,钱苏翔,等.强干扰下复杂系统的独立源识别方法[J].机械工程学报,2006,42(9):149-153.
[2] Li Y Q,Amari S-I,Cichocki A,et al.Underdetermined blind source separation based on sparse representation[J].IEEE Transactions on Signal Processing,2006,54(2):423-437.
[3]董绍江,汤宝平,张焱.基于最优匹配跟踪算法的单通道机械信号盲源分离[J].振动工程学报,2012,25(6):724-731.
[4]吕亚平,李志农,曾宇冬.基于Cohen类时频分布的机械故障源分离[J].机床与液压,2010,38(7):134-137.
[5]吕亚平.基于时频分析的机械故障源盲分离方法研究[D].郑州:郑州大学,2009.
[6]李志农,刘卫兵,易小兵.基于局域均值分解的机械故障欠定盲源分离方法研究[J].机械工程学报,2011,47(7):97-102.
[7]李志农,吕亚平,韩捷.基于时频分析的机械设备非平稳信号盲分离[J].机械强度,2008,30(3):354-358.
[8] Boashash B,A6ssa-El-Bey A,Al-Sa’d M F.Multisensor time-frequency signal processing MATLAB package:an analysis tool for multichannel non-stationary data[J].SoftwareX,2018.DOI:10.1016/j.softx.2017.12.002.
[9] Boashash B.Time-frequency signal analysis and processing:a comprehensive reference[M].Elsevier Ltd.,2016.
[10]Qiu H,Lee J,Lin J,et al.Wavelet filter-based weak signature detection method and its application on rolling element bearing prognostics[J].Journal of Sound and Vibration,2006,289(4):1066-1090.
[2] Li Y Q,Amari S-I,Cichocki A,et al.Underdetermined blind source separation based on sparse representation[J].IEEE Transactions on Signal Processing,2006,54(2):423-437.
[3]董绍江,汤宝平,张焱.基于最优匹配跟踪算法的单通道机械信号盲源分离[J].振动工程学报,2012,25(6):724-731.
[4]吕亚平,李志农,曾宇冬.基于Cohen类时频分布的机械故障源分离[J].机床与液压,2010,38(7):134-137.
[5]吕亚平.基于时频分析的机械故障源盲分离方法研究[D].郑州:郑州大学,2009.
[6]李志农,刘卫兵,易小兵.基于局域均值分解的机械故障欠定盲源分离方法研究[J].机械工程学报,2011,47(7):97-102.
[7]李志农,吕亚平,韩捷.基于时频分析的机械设备非平稳信号盲分离[J].机械强度,2008,30(3):354-358.
[8] Boashash B,A6ssa-El-Bey A,Al-Sa’d M F.Multisensor time-frequency signal processing MATLAB package:an analysis tool for multichannel non-stationary data[J].SoftwareX,2018.DOI:10.1016/j.softx.2017.12.002.
[9] Boashash B.Time-frequency signal analysis and processing:a comprehensive reference[M].Elsevier Ltd.,2016.
[10]Qiu H,Lee J,Lin J,et al.Wavelet filter-based weak signature detection method and its application on rolling element bearing prognostics[J].Journal of Sound and Vibration,2006,289(4):1066-1090.