针对风力机齿轮箱振动响应信号具有强非线性及非平稳性的特点,考虑平均幅值对平均谱负熵时频域成分权重自适应调节,提出连续改进平均谱负熵方法（ICASN）以体现信号细节复杂度特征,并将ICASN引入经验小波变换（EWT）,替代傅里叶谱作为频带划分依据。采用ICASN-EWT分解振动信号,基于改进平均谱负熵筛选特征分量,剔除信号冗余与噪声影响。分析各敏感分量分形特征并构建高维特征集,采用流形学习进行维数约简,并结合分形高斯噪声改进灰狼算法优化支持向量机关键参数,将降维后的向量集输入优化支持向量机进行故障识别与诊断,准确率高达100%。
Abstract
Since the vibration response signal of wind turbine gearbox is highly nonlinear and non-stationary, under the premise of considering the adaptive adjustment of the average amplitude to the average spectral negative entropy of time and frequency domain component weight, the improved continuous average spectral negentropy （ICASN） method is proposed to reflect the detail complexity characteristics of signals. Moreover, ICASN is introduced into Empirical Wavelet Transform （EWT） to replace Fourier spectrum as the basis of frequency band division. According to ICASN-EWT decomposition of vibration signals, the feature components are screened based on Improved Average Spectral Negentropy （IASN） to eliminate signal redundancy and noise influence. Then, the fractal characteristics of each sensitive component are analyzed and the high dimensional feature set is constructed. Meanwhile, Manifold Learning （ML） is used for dimension reduction. Moreover, take fractal Gaussian Noise Grey Wolf Optimizer （FGNGWO） to optimize the key parameters of Support Vector Machine （SVM）. The vector set after dimensionality reduction is input into the optimized support vector machine for fault identification and diagnosis, and the accuracy is up to 100%.
关键词
风力机 /
齿轮箱 /
故障检测 /
支持向量机 /
经验小波变换 /
连续改进平均谱负熵 /
分形高斯噪声改进灰狼算法
Key words
wind turbines /
gearbox /
fault detection /
support vector machines /
empirical wavelet transform /
improved continuous average spectral negentropy /
fractal Gaussian noise grey wolf optimizer
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 全球能源互联网发展合作组织. 中国2030年前碳达峰研究报告[R]. 北京: 中国碳达峰碳中和成果发布暨研讨会, 2021.
Global Energy Internet Development Cooperation Organization. Research report on carbon peak before 2030 in China[R]. Beijing: Publication and Seminar on Carbon Neutrality in China, 2021.
[2] 全球能源互联网发展合作组织. 中国2060年前碳中和研究报告[R]. 北京: 中国碳达峰碳中和成果发布暨研讨会, 2021.
Global Energy Internet Development Cooperation Organization. Research report on carbon neutrality before 2060 in China[R]. Beijing: Publication and Seminar on Carbon Neutrality in China, 2021.
[3] GWEC. Global wind report-annual market update 2021[R]. Washington: Global Wind Energy Council, 2021.
[4] ARTIGAO E, MARTÍN-MARTÍNEZ A, HONRUBIA-ESCRIBANO A, et al. Wind turbine reliability: a comprehensive review towards effective condition monitoring development[J]. Applied energy, 2018, 228: 1569-1583.
[5] QU J X, ZHANG Z S, GONG T.A novel intelligent method for mechanical fault diagnosis based on dual-tree complex wavelet packet transform and multiple classifier fusion[J]. Neurocomputing, 2016, 171: 837-853.
[6] STETCO A, DINMOHAMMADI F, ZHAO X Y, et al.Machine learning methods for wind turbine condition monitoring: a review[J]. Renewable energy, 2019, 133: 620-635.
[7] 隋文涛, 张丹, WANG W.基于EMD和MKD的滚动轴承故障诊断方法[J]. 振动与冲击, 2015, 34(9): 55-59, 64.
SUI W T, ZHANG D, WANG W.Fault diagnosis of rolling element bearings based on EMD and MKD[J]. Journal of vibration and shock, 2015, 34(9): 55-59, 64.
[8] 王志坚, 韩振南, 刘邱祖, 等. 基于MED-EEMD的滚动轴承微弱故障特征提取[J]. 农业工程学报, 2014,30(23): 70-78.
WANG Z J, HAN Z N, LIU Q Z, et al.Weak fault dragnosis for rolling element bearing based on MED-EEMD[J]. Transactions of the Chinese Society of Agricultural Engineering, 2014,30(23): 70-78.
[9] GILLES J.Empirical wavelet transform[J]. IEEE transactions on signal processing, 2013, 61(16): 3999-4010.
[10] BREGLER C, OMOHUNDRO S M.Nonlinear manifold learning for visual speech recognition[C]//Proceedings of the Fifth International Conference on Computer Vision, IEEE, Springer, Berlin, Heidelberg, Germany, 1995: 583-588.
[11] SCHÖLKOPF B, SMOLA A, MÜLLER K R. Kernel principal component analysis[C]//International Conference on Artificial Neural Networks, Springer, Berlin, Heidelberg, Germany, 1997: 583-588.
[12] ANTONI J.The infogram: entropic evidence of the signature of repetitive transients[J]. Mechanical systems and signal processing, 2016, 74: 73-94.
[13] FENG Z P, MA H Q, ZUO M J.Spectral negentropy based sidebands and demodulation analysis for planet bearing fault diagnosis[J]. Journal of sound and vibration, 2017, 410: 124-150.
[14] UGWIRI M A, CARRATU M, PACIELLO V, et al.Benefits of enhanced techniques combining negentropy, spectral correlation and kurtogram for bearing fault diagnosis[J]. Measurement, 2021, 185: 110013.
[15] GILLES J.Empirical wavelet transform[J]. IEEE transactions on signal processing, 2013, 61(16): 3999-4010.
[16] 徐敏, 黄昭毅, 虞和济. 设备故障诊断手册[M]. 西安: 西安交通大学出版社, 2011: 12-19.
XU M, HUANG Z Y, YU H J.Equipment fault diagnosis manual[M]. Xi’an: Xi’an Jiaotong University Press, 2011: 12-19.
[17] ROLO-NARANJO A, MONTESINO-OTERO M E. A method for the correlation dimension estimation for on-line condition monitoring of large rotating machinery[J]. Mechanical systems and signal processing, 2005, 19(5): 939-954.
[18] GRASSBERGER P, PROCACCIA I.Characterization of strange attractors[J]. Physical review letters, 1983, 50(5): 346-349.
[19] 哈特·戴维, 华南理工大学分形课题组, 译. 重分形:理论及应用[M]. 北京: 科学出版社, 2012.
DAVID H.Refractals: theory and application[M]. Beijing: Science Press, 2012.
[20] MIRJALILI S, MIRJALILI S M, LEWIS A.Grey wolf optimizer[J]. Advances in engineering software, 2014, 69: 46-61.
[21] NREL. Wind turbine gearbox vibration condition monitoring benchmarking datasets[R]. Colorado: National Renewable Energy Laboratory, 2012.
[22] ZHANG Z Y, ZHA H Y.Principal manifolds and nonlinear dimension reduction via local tangent space alignment[J]. SIAM journal on entific computing, 2003, 26(1): 313-338.
[23] ROWEIS S T, SAUL L K.Nonlinear dimensionality reduction by locally linear embedding[J]. Science, 2000, 290(5500): 2323-2326.
[24] BELKIN M, NIYOGI P.Laplacian eigenmaps for dimensionality reduction and data representation[J]. Neural computation, 2003, 15(6): 1373-1396.
[25] TENENBAUM J B, SILVA V D, LANGFORD J C.A global geometric framework for nonlinear dimensionality reduction[J]. Science, 2000, 290(5500): 2319-2323.
基金
国家自然科学基金（51976131; 52006148; 52106262）; 上海“科技创新行动计划”地方院校能力建设项目（19060502200）
PDF(13384 KB)
