基于声信号分析的齿轮故障诊断方法

Method of gear fault diagnosis based on acoustic signal analysis

为了解决齿轮故障诊断中传统的声振信号分析方法容易受到周围设备及环境噪声干扰的问题,提出了一种独立分量分析和自相关分析相结合的齿轮故障诊断方法.首先用独立分量分析分离特征信号和干扰信号,然后用自相关分析提取特征信号中的周期成分.实验结果表明,该方法可以有效地提取在强背景噪声干扰下的齿轮故障特征.

Due to the noise disturbance from surrounding equipments and environment in gear fault diagnosis based on acoustic signal analysis, a novel method combined with independent component analysis and auto-correlation was proposed. Independent component analysis was used to separate characteristic signal and interference signal. And then auto-correlation was used to extract the periodic component of characteristic signal. Experimental results indicate that the proposed approach could reduce strong background noise and extract fault feature form gear acoustic signal.