为了提升轴承故障诊断性能,提出了一种基于鲁棒局部均值分解(RLMD)和Kmeans++的轴承故障诊断方法。利用RLMD方法对轴承振动信号进行分解,得到乘积函数(PF),根据PF分量与原始振动信号的相关程度选择敏感PF分量,叠加敏感PF分量构成重构信号;通过计算原始振动信号和重构信号的时域、频域统计特征形成轴承故障特征集;利用线性判别分析(LDA)提取轴承故障的Fisher特征;通过Kmeans++聚类的方法对故障特征进行聚类,得到各工况轴承的聚类中心;通过计算测试样本与聚类中心之间的汉明贴近度来实现轴承故障诊断。利用含有不同信噪比的仿真轴承故障数据和Paderborn大学轴承数据中心的轴承故障数据评价所提出方法的有效性。结果表明,该方法即使在样本数较少的情况下也能够准确地识别出不同类别和级别的轴承故障。
Abstract
To improve the performance of bearing fault diagnosis, a bearing fault diagnosis method based on Robust Local Mean Decomposition(RLMD)and Kmeans++is proposed.The product functions(PF)are obtained by decomposing the bearing vibration signal using the RLMD technique.The sensitive PF components are sifted by calculating the correlation coefficients between the PF components and the original vibration signal, and the sensitive PF components are superimposed to form the reconstructed signal.The bearing fault feature set is formed by calculating the time and frequency domain statistical features of the original vibration signal and the reconstructed signal.The Fisher features of bearing failure feature are extracted by linear discriminant analysis(LDA). The fault feature is clustered by the Kmeans++ clustering method and the cluster center of each bearing working condition is got. The bearing fault identification is realized by calculating the Hamming approach degree between the test sample and the cluster center. The simulated bearing data with different signal-to-noise ratios and bearing data from the Paderborn university test bench are used to evaluate the effectiveness of the proposed method.Results show that the proposed method can accurately identify bearing faults with different categories and levels even though the number of training sample is small.
关键词
轴承 /
故障诊断 /
鲁棒局部均值分解 /
线性判别分析 /
Kmeans++ /
汉明贴近度
{{custom_keyword}} /
Key words
Bearing /
Fault diagnosis /
Robust local mean decomposition(RLMD) /
Linear discriminant analysis(LDA) /
Kmeans++ /
Hamming approach degree
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 邢蓉, 高丙朋, 侯培浩, 等.基于MSCNN与STFT的滚动轴承故障诊断研究[J].机械传动, 2020, 44(7):41-45.XING Rong, GAO Bingpeng, HOU Peihao, et al.Research on fault diagnosis of rolling bearing based on MSCNN and STFT[J].Journal of Mechanical Transmission, 2020, 44(7):41-45.
[2] YU G, LIN T R, WANG Z H, et al.Time-reassigned multisynch-rosqueezing transform for bearing fault diagnosis of rotating machin-ery[J].IEEE Transactions on Industrial Electronics, 2021, 68(2):1486-1496.
[3] HE D, CAO H R, WANG S B, et al.Time-reassigned synchrosqueez-ing transform:the algorithm and its applications in mechanical sig-nal processing[J].Mechanical Systems and Signal Processing, 2019, 117:255-279.
[4] SUN W, YANG G A, CHEN Q, et al.Fault diagnosis of rolling bear-ing based on wavelet transform and envelope spectrum correlation[J].Journal of Vibration and Control, 2013, 19(6):924-941.
[5] LI H, ZHENG H Q, TANG L W.Hilbert-huang transform and its ap-plication in gear faults diagnosis[J].Key Engineering Materials, 2005, 524:655-660.
[6] SMITH J S.The local mean decomposition and its application to EEG perception data[J].Journal of the Royal Society Interface, 2005, 2(5):443-454.
[7] LIU Z L, JIN Y Q, ZUO M J, et al.Time-frequency representation based on robust local mean decomposition for multicomponent AM-FM signal analysis[J].Mechanical Systems and Signal Processing, 2017, 95:468-487.
[8] 樊帆, 徐亚兵.基于半监督聚元自组织映射的齿轮早期故障检测[J].机械传动, 2011, 35(11):66-70.FAN Fan, XU Yabing.Gear early fault detection based on semi-su-pervised convergent self-organizing map[J].Journal of Mechanical Transmission, 2011, 35(11):66-70.
[9] ARTHUR D, VASSILVITSKII S.K-means++:the advantages of careful seeding[C].Proceedings of the Eighteenth Annual ACM-SI-AM Symposium on Discrete Algorithms, 2007:1027-1035.
[10] JIANG Y, LIU S, PENG L, et al.A novel wind speed prediction meth-od based on robust local mean decomposition, group method of data handling and conditional kernel density estimation[J].Energy Con-version and Management, 2019, 200(15):1-15.
[11] LIN T R, KIM E, TAN A C C.A practical signal processing approach for condition monitoring of low speed machinery using peak-hold-down-sample algorithm[J].Mechanical Systems and Signal Process-ing, 2013, 36(2):256-270.
[12] Paderborn University Bearing Data Center. Bearing data[DB/OL].[2016-12-31].https://mb.uni-paderborn.de/kat/forschung/datacen-ter/bearing-datacenter.
[13] LESSMEIER C, KIMOTHO J, ZIMMER D, et al.Condition monitor-ing of bearing damage in electromechanical drive systems by using motor current signals of electric motors:A benchmark data set for da-ta-driven classification[C].European Conference of the Prognostics and Health Management Society, 2016:1-17.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
基金
青岛理工大学新工科研究与实践项目(F2018-119)
{{custom_fund}}