基于双谱的滚动轴承局部损伤故障诊断方法研究

A Diagnosis Method Study Based on Bispectrum for Local Fault Rolling Bearing

针对故障滚动轴承振动信号的耦合特征,将双谱和双相干谱分析方法用于滚动轴承故障特征的提取,给出了外圈、内圈、滚动体局部损伤三种不同状态下轴承振动信号的双谱及双相干谱图,基于双谱形成了故障特征向量。试验表明,对于损伤元件不同的轴承故障,双谱特征图谱存在明显的差异,而且双谱比双相干谱更有利于故障特征的提取,为局部损伤轴承的故障诊断提供有效的方法。

A method based on the bispectrum and the bicoherence analysis was introduced into rolling bearing fault feature extraction, according to the nonlinear coupling properties of fault bearing vibration signals. The bispectrum and the bicoherence of outer-race local fault bearing signals, inner-race local fault bearing signals and ball local fault bearing signals were extracted, and fault feature vectors were set up based on bispectrum. Experimental results show that the bispectra patterns of different component fault are different, the bispectrum has an advantage over the bicoherence for feature extraction, and the fault position recognition methodology based on bispectrum is effective.