基于MED和WMSDL的滚动轴承内圈故障特征诊断

Fault Characteristics Diagnosis of Rolling Bearing Inner Ring Based on MED and WMSDL

针对滚动轴承早期内圈故障特征较为微弱,并伴随环境噪声的干扰,微弱的故障特征信息易被环境噪声所淹没的问题,课题组提出基于最小熵解卷积(MED)和加权多尺度字典学习(WMSDL)的滚动轴承早期故障诊断方法。课题组通过设置一个滤波器使故障特征信号峭度最大实现解卷积,利用WMSDL对解卷积后的信号稀疏分解后进行平方包络解调突出内圈故障特征频率。仿真分析和实例分析结果表明:解卷积后信号的信噪比明显提高,内圈冲击成分明显增强。课题组的研究可有效提取滚动轴承故障特征频率。

Aiming at the problem that the early fault characteristics of the inner ring of rolling bearing were weak and accompanied by the interference of environmental noise, and the weak fault characteristics were easily drowned by the environmental noise, a rolling bearing early fault diagnosis method based on minimum entropy deconvolution(MED) and weighted multi-scale dictionary learning(WMSDL) was proposed. Deconvolution was realized by setting a filter to maximize the kurtosis of fault characteristic signal. The fault feature frequency of the inner ring was highlighted by square envelope demodulation after sparse decomposition of the deconvolution signal with WMSDL. The simulation results show that the signal-to-noise ratio and the impact component of inner ring are enhanced obviously after deconvolution. The proposed method can extract fault frequency of rolling bearing effectively.