机器故障功率谱诊断的进一步探讨

Improved Study for Power Spectrum Diagnosis of Machine Fault

机器故障往往反映为传感信号功率谱的特征变化,为了在复杂机械工况的在线监视与故障诊断中快速有效地捕捉住功率谱的变化特征,本文进一步探讨了表征信号功率谱特征变化的特征量,谱形扩展系数及其快速算法。与过去的几种典型特征量比较,谱形扩展系数具有适应性强且计算更简便等特点。

Generally, machine fault can be detected by the characteristic changes of the power spectrum of a sensing signal.How to recognize and quantitatively translate these signal changes is, therefore, of great importance.A specific variable defined as the expanding coefficient of spectrum which describes the characteristic changes of power spectrum of the signal,is introduced and the fast algorithm of the variable is discussed.It is highly noise proof with good adaptability and requiring less computing time.