谐波窗分解样本熵与灰色关联度在转子故障识别中的应用

Harmonic Window Decomposition Sample Entropy and Grey Relation Degree in Rotor Fault Recognition

针对实测转子振动信号的非平稳特性和在现实条件中难以获得大量典型故障样本的问题,提出一种基于谐波窗分解(harmonic window decomposition,HWD)、样本熵与灰色关联度相结合的故障识别方法。首先,为了降低噪声的影响,引入循环统计学的思想对传统形态滤波方法进行改进,定义了顺序形态滤波器,并结合实际选用最简单的直线结构元素,对实测转子振动信号进行顺序形态滤波降噪预处理。然后,采用不分层分析的HWD来提取包含转子典型故障信息的6个特征频带,运用非线性动力学参数样本熵作为特征,计算转子正常、不平衡、不对中、油膜涡动、油膜振荡等5种工况的样本熵。最后,由于灰色关联度分析对小样本模式识别具有良好的分类效果,以特征频带的样本熵为元素构造特征向量,通过计算不同振动信号的灰色关联度来判断转子的工作状态和故障类型。试验分析结果表明,所提的方法能够有效地应用于转子系统的故障诊断。

Considering the non-stationary features of vibration signal from rotor and the difficulty to obtain enough fault samples in practice,a novel comprehensive fault recognition method was presented based on the harmonic window decomposition（HWD）,sample entropy and grey relation degree.Firstly,the idea of circle statistics was introduced to improve the shortcoming of traditional morphological filter and the rank-order morphological filter was defined;then the line structure element was selected for rank-order morphological filter to denoise the original signal.Secondly,the six feature frequency bands which contain the typical fault information were extracted by harmonic window decomposition that need not decomposition;then the nonlinear dynamic parameter sample entropy was used as a feature and calculated for five rotor conditions.Finally,due to the grey relation degree has good performance in small-sample classification,these sample entropies could serve as the feature vectors,then the grey relation degree of different vibration signals was calculated to identify the fault pattern and condition.Practical results show that this method can identify rotor fault patterns effectively.