摘要
针对滚动轴承故障信号特征难以提取与故障诊断效率较低问题,引入集合经验模态分解(EEMD)对Hilbert-Huang变换(HHT)进行改进,将改进的HHT结合拉普拉斯得分(Laplacian score,LS)进行轴承故障特征提取,并利用遗传算法(GA)优化支持向量机(SVM)分类参数,将其应用于滚动轴承振动信号故障状态识别中。首先,利用相关系数筛选EEMD分解后的IMF分量,计算IMF分量的Hilbert边际谱能量与Lempel-Ziv复杂度构成轴承高维特征向量;其次,运用LS得分对高维特征向量进行数据降维;最后,用GA-SVM对轴承不同故障状态进行识别。通过轴承不同状态下的试验数据验证本文方法,结果表明所提方法能够有效识别轴承不同故障状态。
In order to solve the difficulty to extract the vibration signal characteristics and the problems of low efficiency of fault diagnosis for rolling bearings,the ensemble empirical mode decomposition(EEMD)was introduced to improve the Hilbert-Huang transform(HHT).The improved HHT was then combined with the Laplacian score(LS)for bearing fault feature extraction,and the genetic algorithm(GA)was applied to optimize the parameters of support vector machine(SVM),and for the fault state recognition of rolling bearing vibration signal.Firstly,the correlation coefficient was used to filter the IMF components after EEMD decomposition,and the Hilbert marginal spectral energy and Lempel-Ziv complexity of the IMF components were calculated to form the high-dimensional feature vector.Secondly,the LS was utilized to reduce the dimension of the high-dimensional feature vector.Finally,the GA-SVM was applied to identify different fault conditions of the bearings.The test data of different bearing states verified the proposed method.The results show that the proposed method can effectively identify different fault states of the bearings.
作者
王圣杰
殷红
彭珍瑞
WANG Shengjie;YIN Hong;PENG Zhenrui(School of Mechanical Engineering,Lanzhou Jiaotong University,Lanzhou 730070,China)
出处
《噪声与振动控制》
CSCD
北大核心
2021年第1期89-94,107,共7页
Noise and Vibration Control
基金
甘肃省自然科学基金资助项目(17JR5RA102)
甘肃省高校协同创新团队资助项目(2018C-12)
兰州市人才创新创业资助项目(2017-RC-66)。
作者简介
王圣杰(1994-),男,甘肃省民勤县人,硕士研究生,主要研究方向为齿轮箱故障诊断;通信作者:彭珍瑞,男,博士生导师。E-mail:pzrui@163.com。
