基于全矢谱时间固有尺度分解和独立分量分析盲源分离降噪的滚动轴承故障特征提取

结合盲源分离技术和全矢谱技术的各自优势，提出一种同源双通道信噪盲源分离法。首先采用时间固有尺度分解（ITD）和独立分量分析（ICA）相结合的分析法降噪，对同源双通道的轴承信号进行ITD分解，根据相关系数准则将分解得到的PRC分量进行重组作为ICA输入矩阵，再采用FastICA解混，实现故障信号与噪声信号的分离；其次采用全矢谱技术对信噪分离降噪后的双通道有效分量信号进行全矢信息融合，做全矢谱分析。滚动轴承故障实验对比分析表明了该方法的有效性。

Fault Feature Extraction of Rolling Bearings Based on Noises Reduced by Full Vector Spectrum ITD-ICA Blind Source Separation

The advantages of blind source separation and full vector spectrum technology were combined, and a method of signal noise ratio(SNR) of blind source separation for two channels was presented. Firstly, the ITD and ICA method were combined to reduce noises. The vibration signals of two channels were decomposed by ITD. The proper rotation components(PRC) were decomposed by correlation coefficient criterion restructuring as the inputs of ICA matrix. FastICA was used to realize the separation of fault signals and noise signals. Secondly, the full vector spectrum technology was used to carry out the full vector information fusion of the two channel fault signals and full vector spectrum analysis was completed after reducing noises. The comparison of experiments of rolling bearing fault analyses indicates the effectiveness of the method.