论述基于时频分析的机械故障源盲分离方法,综述时频域盲分离的国内外研究现状。在此基础上,重点论述三种典型的时频域机械故障源分离(Cohen类时频域盲分离、分数Fourier变换域盲分离、EMD域盲分离)方法的技术现状,并进行比较。指出该领域内值得进一步研究的问题。
The machine fault blind source separation method in time-frequency domain, which was named TF-BSS, was introduced. Its state of the art of current research was reviewed. The three typical algorithms, namely Cohen blind source separation, fractional Fourier transform blind source separation and empirical mode decomposition blind source separation, were discussed. Its further research problems were put forward.