融合小波分解与时频分析的单通道振动信号盲分离方法

针对单通道振动信号盲源分离是一个病态问题，且传统的振动信号盲源分离方法往往忽略信号的非平稳性的问题,提出了一种融合小波分解与时频分析的单通道振动信号盲源分离方法。首先利用小波分解与重构将单通道信号转化为多通道信号，解决了盲源分离的欠定问题;然后利用基于时频分析的盲源分离算法分析非平稳信号，得到源信号的估计信号,实现了非平稳信号盲源分离。仿真和实验结果表明，该方法可以有效地解决单通道非平稳振动信号的盲源分离问题。

Single-channel Vibration Signal Separation by Combining Wavelet Decomposition with Time-frequency Analysis

Single-channel  mechanical  vibration  signal-separation  is  an  ill-conditioned  problem,and  in traditional methods, the blind source separation of vibration signals often ignores the nonstationarity.For this reason,a method on single-channel   vibration  signal  separation  was proposed based  on  wavelet decomposition and TFA.The method firstly used  wavelet decomposition and reconstruction   to make the single-channel signals into multi-channel signals,solving the problem of   underdetermined blind source separation;Secondly,based on time-frequency analysis   BSS was used  to effectively analyze the non-stationary signals, then the estimation source signals were obtained,achieving blind source separation of non-stationary signals.Simulation and experimental results vertify the effectiveness of this method, and show the method can solve the problem of BSS of nonstationary single-channel vibration signals.