针对传统的硬阈值奇异值分解降噪法（HSVD）阈值选取主观性较强、自适应性较弱、易丢失信号特征的问题,首先提出一种自适应的硬阈值选取算法;其次,利用一种非等量最优权值收缩的软阈值奇异值分解降噪（SSVD）方法,并结合HSVD,形成一种混合阈值的奇异值分解（SHSVD）降噪方法;最后再结合所提出的一种幅值抑制（AS）算法用于突出信号的故障冲击特征SHSVD-AS。利用该方法对风电传动系统齿轮箱故障信号进行分析,仿真、实测信号的结果均表明,在强噪声环境下,相较于传统的HSVD、VMD-HSVD方法,SHSVD-AS在风电齿轮故障诊断上性能较好。
Abstract
Aiming at the problems of traditional hard threshold singular value decomposition（HSVD） noise reduction have strong subjectivity, weak adaptability and easy to lose signal characteristics, this paper firstly proposes an adaptive hard threshold selection algorithm. Then, a soft-hard threshold singular value decomposition（SHSVD） denoising method is formed by combining an unequal optimal weight shrinkage of soft threshold singular value decomposition （SSVD） denoising method with HSVD. Finally, this paper creates an amplitude suppression（AS） algorithm to highlight the impact characteristics of fault signal denoised by SHSVD, which is SHSVD-AS. This method is used to analyze the gearbox fault signal of wind power transmission system. The test results of simulation and measured signals both indicate that SHSVD-AS has better performance in wind power gear fault diagnosis than traditional HSVD and VMD-HSVD methods under a strong noise enviroment.
关键词
风电机组 /
故障诊断 /
奇异值分解 /
齿轮箱 /
硬阈值 /
软阈值 /
幅值抑制
Key words
wind turbines /
fault diagnosis /
singular value decomposition /
gearbox /
hard threshold /
soft threshold /
amplitude suppression
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] ZHANG L, HU N Q.Fault diagnosis of sun gear based on continuous vibration separation and minimum entropy deconvolution[J]. Measurement, 2019, 141: 332-344.
[2] LI X Y, LI J L, QU Y Z.Semi-supervised gear fault diagnosis using raw vibration signal based on deep learning[J]. Chinese journal of aeronautics, 2020, 33(2): 418-426.
[3] KLEMA V, LAUB A.Singular value decomposition: its computation and some applications[J]. IEEE transactions on automatic control, 1980, 25(2): 164-176.
[4] 汤宝平, 蒋永华, 张详春. 基于形态奇异值分解和经验模态分解的滚动轴承故障特征提取方法[J]. 机械工程学报, 2010, 46(3): 37-48.
TANG B P, JIANG Y H, ZHANG X C.Feature extraction method of rolling bearing fault based on singular value decomposition-morphology filter and empirical mode decomposition[J]. Journal of mechanical engineering, 2010, 46(3): 37-48.
[5] 许午珍, 崔立, 任德余, 等. 基于小波-VMD-Teager能量算子的滚动轴承微弱故障诊断[J]. 上海第二工业大学学报, 2020, 37(3): 200-206.
XU W Z, CUI L, REN D Y, et al.A bearing weak fault diagnosis method based on wavelet-VMD-Teager energy[J]. Journal of Shanghai Second University of Technology, 2020, 37(3): 200-206.
[6] 蒋丽英, 潘宗博, 刘佳鑫. 基于改进VMD-SVD降噪的齿轮箱故障特征提取[J]. 组合机床与自动化加工技术,2021, 33(3): 4-8.
JIANG L Y, PAN Z B, LIU J X.Fault feature extraction of gearbox based on improved VMD-SVD noise reduction[J]. Modular machine tool & automatic manufacturing technique, 2021, 33(3): 4-8.
[7] 赵凯凯, 张柏林, 刘璘, 等. 基于频域奇异值分解的LPI雷达信号降噪[J]. 电讯技术,2015, 55(12): 1407-1412.
ZHAO K K, ZHANG B L, LIU L, et al.The LPI Radar signal denoising based on the frequency-domain singular value decomposition[J]. Telecommunication engineering, 2015, 55(12): 1407-1412.
[8] 李宏, 褚丽鑫, 刘庆强, 等. SG-VMD-SVD的信号去噪方法研究[J]. 吉林大学学报(信息科学版), 2021, 39(2): 158-165.
LI H, CHU L X, LIU Q Q, et al.Study on signal de-noising method of SG-VMD-SVD[J]. Journal of Jilin University(information science edition), 2021, 39(2): 158-165.
[9] 赵学智, 叶邦彦, 陈统坚. 基于变矩阵结构奇异值分解的信号分解算法[J]. 振动·测试与诊断,2018, 38(6): 1097-1102.
ZHAO X Z, YE B Y, CHEN T J.Signal decomposition algorithm based on varying matrix structure singular value decomposition[J]. Journal of vibration, measurement & diagnosis, 2018, 38(6): 1097-1102.
[10] 唐贵基, 李楠楠, 王晓龙. 综合改进奇异谱分解和奇异值分解的齿轮故障特征提取方法[J]. 中国机械工程,2020, 31(24): 2988-2996.
TANG G J, LI N N, WANG X L.Fault feature extraction method for gears based on ISSD and SVD[J]. China mechanical engineering, 2020, 31(24): 2988-2996.
[11] 赵洪山, 郭双伟, 高夺. 基于奇异值分解和变分模态分解的轴承故障特征提取[J]. 振动与冲击, 2016, 35(22): 183-188.
ZHAO H S, GUO S W, GAO D.Fault feature extraction of bearing faults based on singular value decomposition and variational modal decomposition[J]. Journal of vibration and shock, 2016, 35(22): 183-188.
[12] ECKART C, YOUNG G.The approximation of one matrix by another of lower rank[J]. Psychometrika, 1936, 1(3): 211-218.
[13] 赵学智, 叶邦彦, 陈统坚. 频率添加奇异值分解算法及其在故障特征提取中的应用[J]. 机械工程学报,2021,57(10): 10-20.
ZHAO X Z, YE B Y, CHEN T J.Frequency adding singular value decomposition algorithm and its application in fault feature extraction[J]. Journal of mechanical engineering, 2021, 57(10): 10-20.
[14] SCHARF L.The SVD and reduced rank signal processing[J]. Signal processing, 1991, 25(2): 113-133.
[15] SHABALIN A A, NOBEL A B.Reconstruction of a low-rank matrix in the presence of Gaussian noise[J]. Journal of multivariate analysis, 2013, 118(5): 67-76.
[16] BENAYCH-GEORGES F, NADAKUDITI R R.The singular values and vectors of low rank perturbations of large rectangular random matrices[J]. Journal of multivariate analysis, 2021, 111: 120-135.
基金
国家自然科学基金（51875100）; 江苏省重点研发计划（BE2020034）; 江苏省碳达峰碳中和科技创新专项资金（BA2022214）
PDF(1842 KB)
