针对水电机组故障诊断问题，提出了一种基于集合经验模态分解（EEMD），曲线编码（CC）和隐马尔科夫模型（HMM）的故障识别方法。该方法首先利用EEMD处理机组振动信号，得到一系列本征模态函数（IMFs）并计算其标准差（SDs），然后对IMF标准差形成的曲线进行编码构成特征向量。最后将特征向量作为学习样本输入HMM，通过训练得到各状态的HMM。当待测样本输入各状态HMM时，可根据对比各模型输出的对数似然概率值来判断样本所属状态。实验结果表明，该方法能有效提取机组故障特征，识别故障类型，与常规故障识别方法相比，具有较高的准确率。
Abstract
In order to solve the problem of fault diagnosis of hydropower units, a fault identification method based on ensemble empirical mode decomposition (EEMD), curve code (CC) and hidden Markov model (HMM) was proposed. Firstly, EEMD was used to process the vibration signal of the unit to obtain a series of intrinsic mode functions (IMFs) , and the corresponding Standard Deviations (SDs) were calculated. Then, the curve plotted by the SDs was coded into a digital sequence called CC as a fault feature. Finally, the feature vectors were input as learning samples into the HMM to train and obtain the parameters of the HMM of each state. Through comparing the log-probabilities output of various models, the state of the test sample could be judged. The test results showed that the method can effectively extract the fault characteristics from vibration signal of the unit and identify the fault type. At the same time, compared with conventional fault identification methods, it has a higher accuracy.
关键词
水电机组 /
集合经验模态分解 /
曲线趋势编码 /
隐马尔科夫模型 /
故障识别
{{custom_keyword}} /
Key words
hydropower unit /
ensemble empirical mode decomposition /
curve code /
hidden Markov model /
fault identification
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 韩清凯, 于晓光. 基于振动分析的现代机械故障诊断原理及应用[M]. 北京：科学出版社, 2010.
HAN Qing-kai, YU Xiao-guang. The principle and application of modern mechanical fault diagnosis based on vibration analysis[M]. Beijing: Science Press, 2010.
[2] 彭文季, 罗兴锜, 李福松, 等. 基于频谱法和带偏差单元递归神经网络的水电机组振动故障诊[J]. 机械科学与技术, 2006, 25(11):1281-1284.
PENG Wen-ji, LUO Xing-qi, LI Fu-song, et al. Vbration fault diagnosis of hydroelectric generation sets based on spectrun analysis and IRN network[J]. Mechanical Science and Technology, 2006, 25(11):1281-1284.
[3] C. Lu, Y. Wang, M. Ragulskis, and Y. J. Cheng, Fault Diagnosis for Rotating Machinery: A Method based on Image Processing, Plos One, 2016, 11(10):e0164111.
[4] N. E. Huang, R. L. Stever. A new view of nonlinear water waves: the Hilbert Spectrum. Annual Review of Fluid Mechanics, 1999, (31):417-457P.
[5] N. E. Huang. A new method for nonlinear and nonstationary time series analysis:Empirical mode decomposition and Hilbert spectral analysis. Proceedings of SPIE - the International Society for Optical Engineering,2000, (4056)：197-209P.
[6] 王婷. EMD算法研究及其在信号去噪中的应用[D]. 哈尔滨:哈尔滨工程大学, 2010.
WANG Ting. Research on EMD algorithm and its Application in signal denoising[D]. Harbin : Harbin Engineering University, 2010.
[7] Z. WU, and N. E. HUANG, Ensemble Empirical Mode Decomposition: A Noised-assisted Data Analysis Method, Advances in Adaptive Data Analysis, 2009,01(01):1-41.
[8] 刘涵. 水电机组多源信息故障诊断及状态趋势预测方法研究[D]. 武汉：华中科技大学, 2019.
LIU Han. Study on Multiple Information Fault Diagnosis and State Trend Forecasting for Hydroelectric Generator Unit[D]. Wuhan: Huazhong University of Science&Technology, 2019.
[9] 田春花. 隐马尔科夫模型及其在机械故障模式识别中的应用[D]. 北京：华北电力大学, 2007.
TIAN Chun-hua. Hidden Markov Model and Its Application in Mechanical Faults Pattern Recognition[D]. Beijing: North China Electric Power University, 2007.
[10] 蒋文君,胡晓,张培,等.基于EEMD近似熵的水电机组振动信号特征提取[J].水力发电学报,2020,39(06):18-27.
JIANG Wen-jun, HU Xiao, ZHANG Pei, et al. Vibration feature extraction of hydropower unit based on ensemble empirical mode decomposition and approximate entropy[J]. Journal of Hydroelectric Engineering, 2020,39(06):18-27.
[11] Liu D, Xiao ZH, Hu X, Zhang CX, Malik OP. Feature extraction of rotor fault based on EEMD and curve code. Measurement. 2019;135:712-724.
[12] 李军，李佳，张世义，等.采用EEMD算法与互信息法的机械故障诊断方法[J]. 华侨大学学报，2018, 39(1):7-13.
LI Jun, LI Jia, ZHANG Shi-yi, et al. Mechanical Fault Diagnosis Method Using EEMD Algorithm and Mutual Information Method[J]. Journal of Huaqiao University(Natural Science), 2018, 39(1): 7-13. (in Chinese)
[13] 赵道利, 马薇, 梁武科, 等. 水电机组振动故障的信息融合诊断与仿真研究[J]. 中国电机工程学报, 2005(20):137-142.
ZHAO Dao-li, MA Wei, LIANG Wu-ke, et al. Research on data fusion fault diagnosis and simulation of hydroelectric units vibration[J]. Proceedings of The Chinese Society for Electrical Engineering, 2005(20):137-142.
[14] 牛德姣. 面向序列数据分析的分布式存储与计算技术研究[D]. 苏州：江苏大学,2018.
NIU De-jiao, Research on the Distributed Storage and Computing
Technology for Sequential Data Analysis[D]. Suzhou: Jiangsu University, 2018.
[15] 李奕江, 张金萍, 李允公. 基于VMD-HMM的滚动轴承磨损状态识别[J]. 振动与冲击, 2018,37(21):61-67.
LI Yi-jiang, ZHANG Jin-ping, LI Yun-gong. Wear state recognition of rolling bearings based on VMD-HMM[J]. Journal of Vibration and Shock, 2018,37(21):61-67.
[16] 邓丰曼. 一种基于LMD与HMM的刀具磨损故障诊断方法[J]. 机械设计与制造工程, 2019,48(12):111-114.
DENG Feng-man. A tool wear fault diagnosis method based on LMD and HMM[J]. Machine Design and Manufacturing Engineering, 2019,48(12):111-114.
[17] 郑晴晴,傅攀,李威霖.CHMM在滚动轴承故障诊断中的应用研究[J].现代制造工程,2013(12):111-115.
ZHENG Qing-qing, FU Pan, LI Wei-lin. Study on application of CHMM in rolling bearing fault diagnosis[J]. Modern Manufacturing Engineering, 2013(12):111-115.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
PDF(4385 KB)
