独立分量分析技术在设备故障诊断中的应用

Application of ICA on fault diagnosis for running equipments

针对设备振动信号复杂难以分离的特点,提出采用独立分量分析技术对多源振动混合信号进行分离降噪和特征提取。实验结果表明,利用该方法可有效对多源信号进行分离降噪,提取特征信号,从而达到提高故障诊断准确率的目的。

Vibration signals of petrochemical running equipments are complicated, diagnosis feature signal is usually difficult to extract.A new method based on Independent Component Analysis is proposed to realize separation and de-noising for multi-source mixed vibration signals.The experiments showed it is feasible to separate and extract feature signal from multi-source vibration signals.