一种基于参数优化变分模态分解的滚动轴承故障特征提取方法  

A method for rolling bearing fault feature extraction based on parametric optimization VMD

作　　者：郑圆 胡建中[1] 贾民平[1] 许飞云[1] 童清俊 ZHENG Yuan;HU Jianzhong;JIA Minping;XU Feiyun;TONG Qingjun(School of Mechanical Engineering,Southeast University,Nanjing 211189,China)

机构地区：[1]东南大学机械工程学院,南京211189

出　　处：《振动与冲击》2020年第21期195-202,共8页Journal of Vibration and Shock

基　　金：国家自然科学基金(51975117)。

摘　　要：变分模态分解(Variational Mode Decomposition,VMD)已被应用于机械故障诊断领域,惩罚因子对分解结果具有重要的影响,针对实际应用中使用单一惩罚因子分解多振源干扰的信号不能有效提取故障特征的问题,提出一种基于参数优化VMD的滚动轴承故障特征提取方法。根据峭度最大值原则确定分解层数K;通过鲸鱼算法优化选择各模态对应的惩罚因子,实现各模态对应最佳惩罚因子的自适应选择,获得信号的最优模态分解;根据峭度准则筛选分解后的模态分量并进行包络解调,提取轴承故障特征。仿真信号和工程数据的分析结果表明,相对于传统VMD、EEMD和快速谱峭度方法,该方法能够有效提升故障特征提取的敏感性,具有一定的工程应用价值。Aiming at the problem of only using a single penalty factor in traditional variational mode decomposition(VMD) being difficult to extract fault features of rolling bearing in practical application, a method forrolling bearing fault feature extraction based on parametric optimization VMD was proposed. Firstly, the number of decomposition layers K was determined according to the maximum kurtosis criterion. Secondly, the penalty factor corresponding to each mode was optimized with the whale algorithm to realize adaptive selection of each mode’s optimal penalty factor, and obtain a vibration signal’s the optimal mode decomposition. Finally, the kurtosis criterion was used to screen the decomposed modal components, perform envelope demodulation, and extract bearing fault features. The improved VMD was used to analyze simulated signals and engineering actual data. Results showed that compared with the traditional VMD, EEMD and the fast speed spectral kurtosis method, the proposed method can effectively improve the sensitivity of fault feature extraction;it is valuable in engineering applicationto a certain extent.

关 键 词：滚动轴承 变分模态分解 鲸鱼算法 特征提取 

分 类 号：TH212[机械工程—机械制造及自动化] TH213.3