关键词: 故障诊断, 滚动轴承, 变分模态分解, 多尺度排列熵, GK模糊聚类

Abstract: To extract fault features of rolling bearing steadily,a method of fault feature extraction for rolling bearing based on Variational Mode Decomposition (VMD) and Multi-scale Permutation Entropy(MPE) was proposed,and GK fuzzy clustering was used to classify the bearing faults.The vibration signals of rolling bearing were decomposed by VMD into a certain number of modal components with fault feature information.Furthermore,the fault feature of each modal component was quantified by MPE,and the average value of each modal component's MPE was taken as the feature vector.GK fuzzy clustering analysis was used to obtain the standard clustering center of fault samples,and the classification and identification of fault was carried out by Euclid approach degree.The proposed method was applied to the experimental data of rolling bearing,and the classification effect was evaluated by the classification coefficient and the average fuzzy entropy.The comparison with the method of empirical mode decomposition based on MPE and GK fuzzy clustering showed that the proposed method had better classification performance and higher accuracy of fault diagnosis.

Key words: fault diagnosis, rolling bearing, variational mode decomposition, multi-scale permutation, GK fuzzy clustering