A Fault Diagnosis Method Combining with Quadratic VMD Screening, MPE and FCM
-
摘要: 针对单向阀振动信号含有背景噪声,故障特征提取困难和诊断精度不高的问题,提出了二次变分模态分解(二次VMD)、多尺度排列熵(MPE)和模糊C均值聚类(FCM)相结合的故障诊断方法。首先,通过二次VMD对振动信号进行分解,再使用双阈值法筛选得到有用的本征模态函数(IMF)。其次,提取重构信号中具有敏感特性的MPE特征。最后,将故障特征输入至FCM得到聚类中心,并根据海明贴近度对待识别样本进行分类。通过多组对比实验,结果表明二次VMD筛选能有效去除噪声及虚假成分,MPE具有更好的敏感故障特征表征能力。同时,使用FCM对模糊特征进行聚类能够取得比传统支持向量机(SVM)更好的效果。Abstract: In view of the problem that the vibration signal of check valve contains the background noise, difficulty in extracting fault features and low diagnostic precision, a fault diagnosis method combining quadratic VMD (quadratic variational mode decomposition), MPE (multiscale permutation entropy) and FCM (fuzzy C-means clustering) is proposed. Firstly, the vibration signals are decomposed by the quadratic VMD, and then the useful IMF (intrinsic mode functions) are obtained by using the double threshold method. Secondly, the MPE features with sensitive characteristic are extracted from the reconstructed signals. Finally, the fault features are input into the FCM to get the cluster centers, and the samples to be identified are classified according to the Hamming approach degree. Through a number of comparison experiments, the results show that the quadratic VMD screening method can remove the background noise and false components, and the MPE has a better ability to characterize the sensitive features. At the same time, the use of FCM to classify fuzzy features can achieve better results than traditional SVM (support vector machine).
-
表 1 不同模态的中心频率
模态数K 中心频率/Hz 2 575 2 309 - - - - 3 569 2 280 2 982 - - - 4 512 1 087 2 284 2 983 - - 5 512 1 085 2 272 2 789 3 016 - 6 511 1 083 2 129 2 365 2 806 3 018 表 2 VMD筛选得到的IMF及最佳模态数
状态 δk ek IMF K′ NOR 0.4522, 0.1698 0.6144, 0.2682 2, 1 2 IRF 0.167, 0.136, 0.131, 0.130 0.228, 0.223, 0.170, 0.130 5, 4, 3, 7 4 ORF 0.256, 0.222, 0.195, 0.172, 0.112 0.297, 0.253, 0.178, 0.163, 0.067 6, 5, 4, 7, 3 5 REF 0.245, 0.173, 0.129, 0.126 0.316, 0.216, 0.199, 0.158 6, 5, 4, 7 4 表 3 二次VMD筛选得到的IMF
状态 δk ek IMF NOR 0.278 2, 0.117 1 0.863 8, 0.136 2 1, 2 IRF 0.256 0, 0.273 8, 0.136 4 0.400 8, 0.382 0, 0.149 3 4, 3, 2 ORF 0.356 0, 0.332 8, 0.278 1 0.404 7, 0.276 5, 0.316 3 3, 2, 4 REF 0.542 6, 0.176 2 0.714 8, 0.232 7 4, 3 表 4 两种方法的诊断标准
方法 聚类中心 对应状态 VMD筛选提取MPE 0.962 8, 0.962 1, 0.471 8, 0.786 0 IRF 0.244 7, 0.280 6, 0.675 6, 0.742 1 ORF 0.104 2, 0.066 7, 0.887 0, 0.906 1 REF 0.048 9, 0.569 6, 0.671 7, 0.482 9 NOR 二次VMD筛选提取MPE 0.948 4, 0.971 3, 0.521 2, 0.765 2 IRF 0.882 0, 0.778 1, 0.230 8, 0.186 4 NOR 0.202 9, 0.262 2, 0.755 2, 0.727 2 ORF 0.038 8, 0.079 6, 0.909 4, 0.883 4 REF 表 5 VMD筛选提取MPE的诊断结果
表 6 二次VMD筛选提取MPE的诊断结果
表 7 二次VMD筛选得到的IMF
状态 IMF δk ek K′ NC 1, 2, 4, 3 0.611, 0.315, 0.158, 0.117 0.319, 0.308, 0.230, 0.089 4 NK 1, 2 0.350 1, 0.275 1 0.906 5, 0.061 0 2 NM 2, 1, 3 0.674 9, 0.306 5, 0.163 7 0.445 5, 0.403 6, 0.140 5 3 表 8 不同分解方法诊断结果对比
方法 状态 诊断结果 精度 分类系数 模糊熵 NC NK NM 原信号+MPE NC 14 5 1 NK 6 13 1 71.67% 0.642 5 0.633 5 NM 0 4 16 VMD筛选+MPE NC 20 0 0 NK 3 17 0 91.67% 0.850 3 0.365 2 NM 0 2 18 二次VMD筛选+MPE NC 20 0 0 NK 3 17 0 95% 0.881 2 0.259 8 NM 0 0 20 表 9 不同特征量的诊断结果对比
方法 状态 诊断结果 精度 分类系数 模糊熵 NC NK NM VMD+SE NC 20 0 0 NK 5 15 0 85% 0.702 8 0.529 6 NM 0 4 16 VMD+PE NC 20 0 0 NK 1 19 0 88.33% 0.738 1 0.494 6 NM 0 6 14 VMD+MPE NC 20 0 0 NK 4 16 0 90% 0.806 4 0.391 7 NM 0 2 18 表 10 不同诊断方法的诊断结果对比
方法 状态 诊断结果 精度/% NC NK NM 二次VMD筛选+MPE+SVM NC 15 5 0 NK 2 18 0 86.667 NM 1 0 19 二次VMD筛选+MPE+FCM NC 20 0 0 NK 3 17 0 95 NM 0 0 20 -
[1] 段佳雷.基于自相关EMD和快速谱峭度消噪的轴承故障诊断方法研究[J].自动化与仪器仪表, 2017(5):101-103 http://d.old.wanfangdata.com.cn/Periodical/zdhyyqyb201705034Duan J L. Research on fault diagnosis method of bearings based on auto-correlation EMD and fast spectral kurtosis denoising[J]. Automation & Instrumentation, 2017(5):101-103(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/zdhyyqyb201705034 [2] Wu Z H, Huang N E. Ensemble empirical mode decomposition:a noise-assisted data analysis method[J]. Advances in Adaptive Data Analysis, 2009, 1(1):1-41 http://d.old.wanfangdata.com.cn/Periodical/dianzixb201305033 [3] Yeh J R, Shieh J S, Huang N E. Complementary ensemble empirical mode decomposition:a novel noise enhanced data analysis method[J]. Advances in Adaptive Data Analysis, 2010, 2(2):135-156 doi: 10.1142/S1793536910000422 [4] Dragomiretskiy K, Zosso D. Variational mode decomposition[J]. IEEE Transactions on Signal Processing, 2014, 62(3):531-544 doi: 10.1109/TSP.2013.2288675 [5] 马增强, 李亚超, 刘政, 等.基于变分模态分解和Teager能量算子的滚动轴承故障特征提取[J].振动与冲击, 2016, 35(13):134-139 http://d.old.wanfangdata.com.cn/Periodical/zdycj201613022Ma Z Q, Li Y C, Liu Z, et al. Rolling bearings' fault feature extraction based on variational mode decomposition and Teager energy operator[J]. Journal of Vibration and Shock, 2016, 35(13):134-139(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/zdycj201613022 [6] 钱林, 康敏, 傅秀清, 等.基于VMD的自适应形态学在轴承故障诊断中的应用[J].振动与冲击, 2017, 36(3):227-233 http://d.old.wanfangdata.com.cn/Periodical/zdycj201703036Qian L, Kang M, Fu X Q, et al. Application of adaptivemorphology in bearing fault diagnosis based on VMD[J]. Journal of Vibration and Shock, 2017, 36(3):227-233(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/zdycj201703036 [7] 刘尚坤, 唐贵基, 王晓龙, 等.基于改进变分模态分解的旋转机械故障时频分析方法[J].振动工程学报, 2016, 29(6):1119-1126 http://d.old.wanfangdata.com.cn/Periodical/zdgcxb201606022Liu S K, Tang G J, Wang X L, et al. Time frequency analysis method for rotary mechanical fault based on improved variational mode decomposition[J]. Journal of Vibration Engineering, 2016, 29(6):1119-1126(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/zdgcxb201606022 [8] 唐贵基, 王晓龙.参数优化变分模态分解方法在滚动轴承早期故障诊断中的应用[J].西安交通大学学报, 2015, 49(5):73-81 http://d.old.wanfangdata.com.cn/Periodical/xajtdxxb201505012Tang G J, Wang X L. Parameter optimized variational mode decomposition method with application to incipient fault diagnosis of rolling bearing[J]. Journal of Xi'an Jiaotong University, 2015, 49(5):73-81(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/xajtdxxb201505012 [9] 朱永利, 贾亚飞, 王刘旺, 等.基于改进变分模态分解和Hilbert变换的变压器局部放电信号特征提取及分类[J].电工技术学报, 2017, 32(9):221-235 http://d.old.wanfangdata.com.cn/Periodical/dgjsxb201709025Zhu Y L, Jia Y F, Wang L W, et al. Feature extraction and classification on partial discharge signals of power transformers based on improved variational mode decomposition and Hilbert transform[J]. Transactions of China Electrotechnical Society, 2017, 32(9):221-235(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/dgjsxb201709025 [10] Wang P, Wang T Y. Energy weighting method and its application to fault diagnosis of rolling bearing[J]. Journal of Vibroengineering, 2017, 19(1):223-236 doi: 10.21595/jve.2016.17338 [11] Bandt C, Pompe B. Permutation entropy:a natural complexity measure for time series[J]. Physical Review Letters, 2002, 88(17):174102 doi: 10.1103/PhysRevLett.88.174102 [12] 杜萌, 金宁德, 高忠科, 等.油水两相流水包油流型多尺度排列熵分析[J].物理学报, 2012, 61(23):230507 doi: 10.7498/aps.61.230507Du M, Jin N D, Gao Z K, et al. Multiscale permutation entropy analysis of oil-in-water type two-phase flow pattern[J]. Acta Physica Sinica, 2012, 61(23):230507(in Chinese) doi: 10.7498/aps.61.230507 [13] 郑近德, 程军圣, 杨宇, 等.多尺度排列熵及其在滚动轴承故障诊断中的应用[J].中国机械工程, 2013, 24(19):2641-2646 doi: 10.3969/j.issn.1004-132X.2013.19.017Zheng J D, Cheng J S, Yang Y, et al. Multi-scale permutation entropy and its applications to rolling bearing fault diagnosis[J]. China Mechanical Engineering, 2013, 24(19):2641-2646(in Chinese) doi: 10.3969/j.issn.1004-132X.2013.19.017 [14] 王军辉, 贾嵘, 谭泊, 等.基于EEMD和模糊C均值聚类的风电机组齿轮箱故障诊断[J].太阳能学报, 2015, 36(2):319-324 doi: 10.3969/j.issn.0254-0096.2015.02.009Wang J H, Jia R, Tan B, et al. Fault diagnosis of wind turbine's gearbox based on EEMD and fuzzy C means clustering[J]. Acta Energiae Solaris Sinica, 2015, 36(2):319-324(in Chinese) doi: 10.3969/j.issn.0254-0096.2015.02.009 [15] Ye Z, Cai G Q, Yang J W, et al. Fault diagnosis of railway rolling bearing based on wavelet analysis and FCM[J]. International Journal of Digital Content Technology and its Applications, 2011, 5(3):47-58 doi: 10.4156/jdcta.vol5.issue3.5 [16] Dragomiretskiy K, Zosso D. Variational mode decomposition[J]. IEEE Transactions on Signal Processing, 2014, 62(3):531-544 doi: 10.1109/TSP.2013.2288675 [17] 刘长良, 武英杰, 甄成刚.基于变分模态分解和模糊C均值聚类的滚动轴承故障诊断[J].中国电机工程学报, 2015, 35(13):3358-3365 http://d.old.wanfangdata.com.cn/Periodical/zgdjgcxb201513020Liu C L, Wu Y J, Zhen C G. Rolling bearing fault diagnosis based on variational mode decomposition and fuzzy C means clustering[J]. Proceedings of the CSEE, 2015, 35(13):3358-3365(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/zgdjgcxb201513020 [18] Battiti R. Using mutual information for selecting features in supervised neural net learning[J]. IEEE Transactions on Neural Networks, 1994, 5(4):537-550 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=a6b63bc9fd1ba7e0a2114612ca6f5623 [19] Aziz W, Arif M. Multiscale permutation entropy of physiological time series[C]//2005 Pakistan Section Multitopic Conference. Karachi, Pakistan: IEEE, 2007: 1-6 [20] Bandt C, Pompe B. Permutation entropy:a natural complexity measure for time series[J]. Physical Review Letters, 2002, 88(17):174102 doi: 10.1103/PhysRevLett.88.174102 [21] 郑直, 姜万录, 胡浩松, 等.基于EEMD形态谱和KFCM聚类集成的滚动轴承故障诊断方法研究[J].振动工程学报, 2015, 28(2):324-330 http://d.old.wanfangdata.com.cn/Periodical/zdgcxb201502020Zheng Z, Jiang W L, Hu H S, et al. Research on rolling bearings fault diagnosis method based on EEMD morphological spectrum and kernel fuzzy C-means clustering[J]. Journal of Vibration Engineering, 2015, 28(2):324-330(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/zdgcxb201502020 [22] Cannon R L, Dave J V, Bezdek J C. Efficient implementation of the fuzzy C-means clustering algorithms[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1986, PAMI-8(2):248-255 doi: 10.1109/TPAMI.1986.4767778