多尺度熵在转子故障诊断中的应用

Rotor Fault Diagnosis Based on Multiscale Entropy

为解决从信号中提取故障特征难的问题,介绍了一种新的信号故障特征提取方法——多尺度熵(multi-scaleentropy,简称MSE),并将样本熵和多尺度熵分别应用于转子故障信号复杂性的度量,以提取故障特征。试验数据分析表明,与样本熵相比,多尺度熵更能有效地实现转子故障类型的诊断。

A new method,multiscale entropy(MSE) is introduced,which can extract the fault feature from the signals.The concepts of sample entropy and multiscale entropy are introduced firstly,and then they are used to measure the complexity of rotor signals.By analyzing rotor experimental data,the results show that,compared with sample entropy,multiscale entropy can distinguish the fault categories of rotor more effectively.