基于APIT-SA-MEMD和FLLE的齿轮故障识别  

Gear Fault Diagnosis Based on APIT-SA-MEMD and FLLE

作　　者：冀浩非 刘慧玲[2] 董加强 Ji Haofei;Liu Huiling;Dong Jiaqiang(School of Mechanical and Electrical Engineering,Jinzhong Vocational and Technical College,Jinzhong 030600,China;Department of Machinery,Jinzhong College,Jinzhong 030600,China;School of Mechanical Engineering,Hefei University of Technology,Hefei 230002,China)

机构地区：[1]晋中职业技术学院机电工程学院,山西晋中030600 [2]晋中学院机械系,山西晋中030600 [3]合肥工业大学机械工程学院,安徽合肥230002

出　　处：《机械传动》2022年第11期161-169,共9页Journal of Mechanical Transmission

摘　　要：齿轮往往处于恶劣的工作环境,其振动信号具有非线性和非平稳性的特点,研究出适用于齿轮的故障诊断方法具有重要意义。针对这一问题,提出了一种基于自适应投影本质变换正弦辅助多元经验模式分解(APIT-SA-MEMD)和Floyd局部线性嵌入算法(FLLE)的智能故障诊断方法。自适应投影本质变换多元经验模式分解存在模态混叠现象,因此,提出自适应投影本质变换正弦辅助多元经验模式分解来减轻传统经验模式分解存在的模态混叠现象。首先,采用APIT-SA-MEMD方法对齿轮振动信号进行分解,获得能够表征齿轮振动信号的IMF分量;在此基础上,提取所选取IMF分量的时域和频域特征,获得高维特征矩阵;最后,利用FLLE对高维特征矩阵进行降维和聚类分析,实现齿轮故障模式的识别。实验结果表明,提出的方法能够准确识别齿轮的不同故障类型。Gears are often in a harsh working environment,and their vibration signals have the characteris⁃tics of non-linearity and non-stationarity.Therefore,it is of great significance to develop a fault diagnosis meth⁃od suitable for gears.To solve this problem,an intelligent fault diagnosis method based on adaptive projection in⁃trinsically transformation sine-assisted multivariate empirical mode decomposition(APIT-SA-MEMD)and Floyd local linear embedding(FLLE)algorithm is proposed.Multivariate empirical mode decomposition of adap⁃tive projection intrinsically transformation has modal aliasing phenomenon,so APIT-SA-MEMD is proposed to reduce the modal aliasing phenomenon existing in traditional empirical mode decomposition.First,the APIT SA-MEMD method is used to decompose the gear vibration signal,and the IMF component that can characterize the gear vibration signal is obtained.On this basis,the time domain and frequency domain features of the select⁃ed IMF components are extracted to obtain a high-dimensional feature matrix.Finally,FLLE is used to perform dimensionality reduction and clustering analysis on the high-dimensional feature matrix to realize the identifica⁃tion of gear fault modes.Experimental results show that the proposed method can accurately identify different types of gear faults.

关 键 词：齿轮 自适应投影本质变换正弦辅助多元经验模式分解 Floyd局部线性嵌入 降维故障识别 

分 类 号：TH132.41[机械工程—机械制造及自动化]