基于时域优选的VMD-SVM滚动轴承故障分类识别方法研究  

Research on Rolling Bearing Fault Identification and Classification Method Based on VMD-SVM with Optimization of Time-Domain Features

作　　者：来永斌[1] 康曦 汪森辉 LAI Yongbin;KANG Xi;WANG Senhui(School of Mechanical and Electrical Engineering,Anhui University of Science and Technology,Huainan 232001,China)

机构地区：[1]安徽理工大学机电工程学院,安徽淮南232001

出　　处：《西安文理学院学报(自然科学版)》2025年第1期1-7,37,共8页Journal of Xi’an University(Natural Science Edition)

摘　　要：针对滚动轴承早期故障特征提取较难以及时域特征选取存在局限性的问题,提出一种基于时域优选的VMD-SVM滚动轴承故障分类识别方法,采用变分模态分解(VMD)和最小包络熵法将故障信号分解成合适数量的本征模态分量(IMF),用峭度指标和相关性系数筛选出敏感分量,基于40个时域特征指标对敏感分量进行初步特征提取,用相关性分析优选出特征指标进行二次特征提取,提取得到的特征向量输入至支持向量机(SVM),用网格搜索和交叉验证方法确定最佳超参数后进行模型训练以及轴承故障分类识别.分别与经验模态分解(EMD)、集合经验模态分解(EEMD)方法比较,VMD-SVM方法准确率达(99±0.12)%,而EMD-SVM、EEMD-SVM方法准确率分别为(84±0.17)%、(89±0.1)%.结果表明本方法能够较高精度地分类识别不同工况下的滚动轴承状态.Argeting the difficulty of capturing early fault features and the limitations of choosing time-domain features in rolling bearings,a fault identification and classification method based on VMD-SVM with optimization of time-domain features is proposed.After the number of layers was determined by the minimum envelope entropy method,the Intrinsic Modal Function(IMF)is decomposed by Variational Modal Decomposition(VMD),and the kurtosis index and correlation coefficient are used to screen out the sensitive components.A preliminary feature extraction is performed on the sensitive components based on the 40 time-domain feature indexes,and the feature indexes are further optimized with the correlation analysis for secondary feature extraction,which are then input into the Support Vector Machine(SVM).After the best hyperparameters are determined through grid search and cross-validation,model training was conducted,followed by the classification and recognition of bearing faults.Compared to the Empirical Modal Decomposition(EMD)and Ensemble Empirical Modal Decomposition(EEMD)methods,the accuracy of the VMD-SVM method reaches(99±0.12)%,while the accuracy of the EMD-SVM and EEMD-SVM methods are(84±0.17)%and(89±0.1)%,respectively.The results show that this method can classify and identify the rolling bearing condition under different working conditions with high accuracy.

关 键 词：时域优选 变分模态分解 支持向量机 滚动轴承 故障分类识别 

分 类 号：TH133.33[机械工程—机械制造及自动化] TP18[自动化与计算机技术—控制理论与控制工程]