基于EEMD多尺度模糊熵的S700K转辙机故障诊断

Fault Diagnosis of S700K Switch Machine Based on EEMD Multi-scale Fuzzy Entropy

为了更加有效地提取S700K转辙机典型故障下动作功率曲线的故障特征,基于集合经验模态分解和多尺度模糊熵的方法提出一种新的故障诊断方法。首先对S700K转辙机动作功率曲线进行集合经验模态分解,计算获得包含不同时间信息的本征函数;然后对不同时间维度的分量提取模糊熵作为特征参数,由于模糊熵能辨别出信号的细微特征,从而可以获得原始信号多个维度的细微特征作为S700K转辙机不同状态下的特征参数;最后根据灰色关联度分析算法计算待检曲线和各故障曲线特征值之间的灰关联度值,由该值确定S700K转辙机的故障类型。实验结果表明,该诊断模型能显著提高S700K转辙机的故障诊断精度,且具有较好地适应性。

In order to more effectively extract the fault characteristics of the action power curve of the S700K switch machine under typical faults,a novel fault diagnosis method was proposed based on ensemble empirical mode decomposition and multi-scale fuzzy entropy.Firstly,the S700K action power curve was decomposed by ensemble empirical mode decomposition,and the intrinsic function containing different time information was obtained by calculation.Then the fuzzy entropy was extracted from the components of different time dimensions as the characteristic parameter.Because the fuzzy entropy can distinguish the subtle features of the signal,the subtle features of the original signals from multiple dimensions of the switch machine were obtained as the characteristic parameters of S700K switch machine in different states.Finally,according to the grey relational degree analysis algorithm,the grey relational degree value between the curve to be checked and the characteristic value of each fault curve was calculated,which value determined the fault type of S700K.The experimental results show that the fault diagnosis model can significantly improve the accuracy of the fault diagnosis of the S700K switch machine,with better adaptability.