基于时频掩蔽和振动特征的齿轮箱振源分离方法

PDF(4364 KB)

振动与冲击 ›› 2023, Vol. 42 ›› Issue (15) : 73-82.

论文

张博1，王罗2，邹祖冰2，滕伟1，邓友汉2

作者信息 +

Vibration source separation algorithm for gear box based on time-frequency masking and vibration signal features

ZHANG Bo1, WANG Luo2, ZOU Zubing2, TENG Wei1, DENG Youhan2

Author information +

文章历史 +

摘要

针对齿轮箱振动信号中存在多源混合问题，提出基于时频掩蔽和振动特征的齿轮箱振源分离算法。算法的主要步骤为：首先通过短时傅里叶变换将时域信号转换为时频谱进行分析，通过最佳比率掩蔽获取齿轮啮合频率特征，其次采用寻峰算法获取混合信号中所有频率成分，结合故障频域特征进一步分离故障成分，通过理想比率掩蔽获取齿轮的调制特征，最后运用短时傅里叶逆变换恢复源信号。通过风电传动试验台信号和实际风电现场振动信号对算法进行验证，结果表明，在已知齿轮转速情况下，所提出的算法可以有效分离源信号并提取故障成分。

Abstract

A vibration source separation algorithm based on time-frequency domain masking and gear vibration signal characteristics is proposed for the problem of multi-source mixing in gearbox observation signals, which can separate individual vibration source information from the multi-source mixed data. The main steps of the algorithm are as follows: firstly, the time domain signal is converted into time-frequency spectrum for analysis by short-time Fourier transform, the gear mesh frequency characteristics are obtained by optimal ratio masking, secondly, all the frequency components in the mixed signal are obtained by peak-seeking algorithm, the fault characteristics are further separated by combining with the fault frequency domain characteristics, the gear modulation components are obtained by ideal ratio masking, and finally the source signal is recovered by inverse short-time Fourier transform. The algorithm is validated by wind power transmission platform and actual wind power field vibration signals, and the results show that the proposed algorithm can effectively extract the main components of the source signal for a known gear speed.

导出引用

张博1，王罗2，邹祖冰2，滕伟1，邓友汉2. 基于时频掩蔽和振动特征的齿轮箱振源分离方法[J]. 振动与冲击, 2023, 42(15): 73-82

ZHANG Bo1, WANG Luo2, ZOU Zubing2, TENG Wei1, DENG Youhan2. Vibration source separation algorithm for gear box based on time-frequency masking and vibration signal features[J]. Journal of Vibration and Shock, 2023, 42(15): 73-82

参考文献

[1] Herault J, Jutten C. Space or time adaptive signal processing by neural network models[C]. Aip Conference 151 on Neural Networks for Computing. American Institute of Physics, 1986:206-211.
[2] Li Y P, Wang D L. On the optimality of ideal binary time-frequency masks[J]. Speech Communication, 2009, 51(3)：230-239.
[3] Hummersone C, Mason R, Brookes T. Ideal binary mask ratio: a novel assessing binary-mask-based sound source separation algorithms[J]. IEEE Transactions on Audio, Speech, and Language Processing, 2011, 19(7):2039-2045.
[4] Narayanan A, Wang D L. Ideal ratio mask estimation using deep neural networks for robust speech recognition[C]. 2013 IEEE International Coference on Acoustics, Speech and Aignal Processing, 2013, 7092-7096.
[5] Liang S, Liu W J, Jiang W, et al. The optimal ratio time-frequency mask for speech separation in terms of the signal-to-noise ratio[J]. The Joural of the Acoustical Society of America, 2013,134(5).
[6] 朱会杰，王新晴，芮挺，等. 基于移不变稀疏编码的单通道机械信号盲源分离[J]. 振动工程学报，2015, 28(4)：625-632.
ZHU Hui-jie, WANG Xin-qing, RUI Ting, et al. Shift invariant sparse coding for blind source separation of single channel mechanical signal[J]. Journal of vibration engineering, 2015，28(4)：625-632.
[7] Hao Y, Song L, Wang M, et al. Underdetermined Source Separation of Bearing Faults Based on Optimized Intrinsic Characteristic-Scale Decomposition and Local Non-Negative Matrix Factorization[J]. IEEE Access, 2019,7:11427-11435.
[8] 杨广振，荆建平，明阳，等. 利用航空发动机信号特征的振动源盲分离算法[J]. 西安交通大学学报，2017, 51(6):20-27.
YANG Guangzhen, JING Jiangping, MING Yang, et al. Blind vibration sources separation method based on signal feature of aircraft engine[J]. Journal of Xi`an jiaotong university, 2017, 51(6):20-27
[9] 刘鲲鹏，夏均忠，白云川，等. EDRS在滚动轴承振动信号盲源分离中的应用[J]. 振动与冲击，2019, 38(20):106-111.
LIU Kunpeng, XIA Junzhong, BAI Yunchuan, et al. Application of EDRS in blind source separation of rolling element bearing vibration signal[J]. Jourmal of vibration and shock, 2019, 38(20):106-111.
[10] 郝如江，安雪君，史云林. 基于EMD和CICA的单通道盲源分离方法用于齿轮箱混合故障诊断研究[J]. 振动与冲击，2019, 38(8):225-231.
HAO Rujiang, AN Xuejun, SHI Yunlin. Single-channel blind source separation based on EMD and CICA and its application to gearbox multi-fault diagnosis[J]. Jourmal of vibration and shock, 2019, 38(8):225-231.
[11] 李志农，张芬，肖尧先. 基于CVA-ICA的机械故障源动态盲分离方法[J]. 机械工程学报, 2015, 51(12):24-29.
LI Zhinong, ZHANG Fen, XIAO Yaoxian. Dynamic Blind Separation of Mechanical Fault Sources Based on
Canonical Variate Analysis and Independent Component Analysis[J].Journal of mechanical engineering, 2015, 51(12):24-29.
[12] Li G, Tang G, Wang H, et al. Blind source separation of composite bearing vibration signals with low-rank and sparse decomposition[J]. Measurement, 2019,145:323-334
[13] 于刚，周以齐，张卫芬. 一种新的相关机械振源盲分离方法[J]. 振动与冲击, 2016, 35(15): 216-221.
YU Gang, ZHOU Yiqi, ZHANG Weifen. Blind source separation of correlated vibration sources[J]. Jourmal of vibration and shock, 2016, 35(15):216-221.
[14] 赵洪山，李浪，王颖. 一种基于盲源分离和流形学习的风电机组轴承故障特征提取方法[J]. 太阳能学报, 2016, 37(2): 269-275.
ZHAO Hongshan, LI Lang, WANG Ying. Fault feature extraction method of wind turbine bearing based on blind source separation and manifold learning[J]. Acta energiae solaris sinica, 2016, 37(2):269-275.
[15] 余路，曲建岭，高峰，等. 基于改进字典学习的单通道振动信号盲源分离算法[J]. 振动与冲击, 2019, 38(1):96-102.
YU Lu, QU Jianling, GAO Feng, et al. Single channel blind source separation of vibration signals based on improved dictionary learning[J]. Jourmal of vibration and shock, 2019, 38(1): 96-102.