摘要
针对轴承振动信号易受噪声影响造成故障特征难以提取的问题,提出一种基于双树复小波(Dual-Tree Complex Wavelet Transform,DTCWT)和最大后验估计(Maximum A Posteriori,MAP)的信号去噪及故障诊断方法。该方法首先对振动信号进行DTCWT分解,获得不同层次的小波系数,有效克服了传统小波分解频率混叠和畸变的缺陷。然后根据各层小波系数中的噪声强度构造MAP软阈值函数,对不同层次的小波系数进行阈值去噪。最后对去噪后的各层小波系数进行DTCWT反变换,将信号分解到不同频带,结合Hilbert包络实现轴承故障诊断。仿真信号去噪分析、轴承复合故障诊断实验及风机轴承微弱故障诊断应用结果表明,该方法能够有效去除噪声,提高信噪比,增强故障特征,提高轴承故障诊断的准确性和实效性。
In order to solve the problem that bearing vibration signal is susceptible to noise and fault features are difficult to extract, a signal denoising and fault diagnosis method based on Dual-Tree Complex Wavelet Transform(DTCWT) and Maximum A Posteriori(MAP) is proposed. Firstly, different levels of wavelet coefficients of the vibration signal are obtained by DTCWT, which effectively overcomes the defects of frequency aliasing and distortion of traditional wavelet decomposition. Then the soft threshold function based on MAP is constructed according to the noise intensity of the different layer wavelet coefficients, and the threshold denoising of the wavelet coefficients is carried out. Finally, the denoised wavelet coefficients of each layer are inversely transformed by DTCWT, and the signals are decomposed into different frequency bands. Combining Hilbert envelope, the bearing fault diagnosis was realized. The results of simulation signal de-noising analysis, bearing compound fault diagnosis experiment and wind turbine bearing weak fault diagnosis application show that the method can effectively remove noise, improve signal-to-noise ratio, enhance fault characteristics, and improve the accuracy and effectiveness of bearing fault diagnosis.
作者
胡永涛
周强
范峥
王国柱
HU Yong-tao;ZHOU Qiang;FAN Zheng;WANG Guo-zhu(School of Electrical Engineering and Automation,Henan Institute of Technology,Xinxiang 453003,China;Weihua Group,Changyuan 453400,China)
出处
《河南工学院学报》
CAS
2020年第1期8-14,共7页
Journal of Henan Institute of Technology
基金
河南省高等学校重点科研项目(19B460001)
河南省科技攻关计划资助项目(202102210061,182102210033,182102210261)。
关键词
振动信号去噪
轴承故障诊断
双树复小波
最大后验估计
vibration signal denoising
bearing fault diagnosis
dual-tree complex wavelet (DTCWT)
maximum a posteriori(MAP)
