基于AVMD与DPC-FCM的旋转机械无监督故障诊断方法
Unsupervised Fault Diagnosis of Rotating Machinery Based on AVMD and DPC-FCM Algorithm
武雅曼 1谌鹏 1张滇 1刘天 1唐剑1
作者信息
- 1. 中国船舶重工集团有限公司第七〇五研究所,西安 710000
- 折叠
摘要
目的 针对旋转机械故障诊断过程中存在故障信号特征提取困难、故障诊断过程有标签数据较少、故障诊断准确率低等问题,提出自适应变分模态分解算法(Adaptive Variational Mode Decomposition,AVMD)与密度峰值算法优化的模糊C均值算法(Clustering by Fast Search and Find of Density Peaks Optimizing Fuzzy C-Means,DPC-FCM)结合的无监督诊断方法.方法 首先,将多尺度排列熵与峭度相结合的综合系数作为适应度函数,对VMD算法的惩罚因子alpha和模态个数K进行参数寻优,提取分解后本征模态函数(Intrinsic Mode Function,IMF)的平均样本熵与平均模糊熵,并输入至聚类算法中.其次,提出利用密度峰值聚类算法确定 FCM 的初始聚类中心,降低聚类结果的随机性.结果 将提出的无监督故障诊断模型应用到滚动轴承试验信号中,实现了准确的故障诊断.结论 AVMD 在故障提取方面具有优越性,同时 DPC 算法可以有效提高FCM算法无监督聚类的准确性,二者结合可以有效实现旋转机械故障的智能分类.
Abstract
In order to solve the problems of difficult feature extraction of rotating machinery fault signals,less label data in the fault diagnosis process,and low fault diagnosis accuracy,the work aims to propose an unsupervised diagnosis method com-bining adaptive variational mode decomposition(AVMD)and clustering by fast search and find of density peaks optimizing fuzzy C-means(DPC-FCM).Firstly,the comprehensive coefficient combining multi-scale permutation entropy and kurtosis was proposed as the fitness function to optimize the penalty factor alpha and modal number K of the VMD algorithm.The average sample entropy and average fuzzy entropy of the decomposed intrinsic mode function were extracted and input into the cluster-ing algorithm.Secondly,to solve the problem that FCM clustering algorithm was easy to fall into the local optimal solution,DPC-FCM was proposed to determine the initial clustering centers of FCM to reduce the randomness of clustering results.Fi-nally,the unsupervised fault diagnosis model was constructed and applied to the rolling bearing testing signals,achieving the accurate fault diagnosis.AVMD has advantages in fault extraction,and DPC algorithm can effectively improve the accuracy of unsupervised clustering of FCM algorithm.The combination of the two can effectively realize intelligent classification of rotat-ing machinery faults.
关键词
变分模态分解算法/模糊C均值/密度峰值聚类/旋转机械/故障诊断Key words
variational mode decomposition/fuzzy C-means/clustering by fast search and find of density peaks/rotating machinery/fault diagnosis引用本文复制引用
出版年
2024
NETL
NSTL
维普
万方数据