作者：宋守许, 汪志全, 蔚辰, 李想

作者单位：合肥工业大学机械工程学院，安徽 合肥 230009

关键词：信号降噪;经验模态分解;奇异值分解降噪;脉冲涡流

摘要：

针对脉冲涡流检测信号非平稳性特点和强背景噪声干扰问题，提出一种经验模态分解（EMD）与改进奇异值分解（SVD）联合的信号降噪方法。首先将信号经EMD分解成一系列的固有模态函数（IMFs），根据各固有模态函数与原信号相关系数，将固有模态函数分为信号主导分量和噪声主导分量。然后使用改进SVD降噪方法对噪声主导分量进行降噪，并与信号主导分量重构得到降噪信号。基于仿真信号和实测信号对新方法的优越性和有效性进行验证，在对实测信号采用该方法处理后，结果表明该方法能将涡流检测信号信噪比提高至30.95 dB，均方误差降低至0.0260，能有效消除强噪声的干扰，为信号特征量的准确提取奠定基础。

Signal denoising method of pulsed eddy current testing signal based on EMD and improved SVD
SONG Shouxu, WANG Zhiquan, WEI Chen, LI Xiang
School of Mechanical Engineering, Hefei University of Technology, Hefei 230009，China
Abstract: Aiming at the non-stationary characteristics of pulse eddy current testing signal and the problem of strong background noise interference, a signal denoising method combining empirical mode decomposition (EMD) and improved singular value decomposition (SVD) was proposed. Firstly, the signal is decomposed into a series of intrinsic mode functions (IMFs) by EMD. According to the correlation coefficient between each intrinsic mode function and the original signal, the intrinsic mode function is divided into signal dominant component and noise dominant component. Then the Improved SVD denoising method is used to denoise the dominant component of the noise, and the denoised signal is reconstructed with the dominant component of the signal. The advantages and effectiveness of the new method are verified based on the simulation signal and the measured signal. The results show that this method can improve the signal-to-noise ratio to 30.95 dB and reduce the mean square error to 0.0260, for which we can say that the method can effectively eliminate the interference of strong noise, and lay a foundation for the accurate extraction of signal characteristics.
Keywords: signal denoising;empirical mode decomposition;singular value decomposition denoising;pulsed eddy current
2022, 48(9):97-104  收稿日期: 2021-05-31;收到修改稿日期: 2021-07-15
基金项目: 国家自然科学基金项目（51975176）
作者简介: 宋守许（1964-），男，安徽霍山县人，教授，博导，博士，主要研究方向为再制造工程
参考文献
[1] YANG G, TIAN G Y, QUE P W, et al. Data fusion algorithm for pulsed eddy current detection[J]. IET Science Measurement and Technology, 2007, 1(6): 312-316
[2] 杨理践, 李春华, 裴磊. 脉冲涡流信号检测与分析[J]. 中国机械工程, 2013, 24(13): 1773-1776+1782
[3] 张智军, 杨博楠, 杜金强. 希尔伯特-黄变换在脉冲涡流信号消噪与识别中的应用[J]. 国防科技大学学报, 2018, 40(3): 61-68
[4] 徐志远, 伍权. 基于维纳滤波和主成分分析的脉冲涡流检测信号降噪方法[J]. 传感技术学报, 2019, 32(3): 411-417
[5] LU Z, WANG F L, CHANG Y Q, et al. Region-adaptive median filter[J]. Journal of System Simulation, 2007, 19(23): 5411-5414
[6] STEVENSON R A, FRYSZ C A. Process for tuning an emi filter to reduce the amount of heat generated in implanted lead wires during medical procedures such as magnetic resonance imaging: US20070083244[P]. 2007-04-12.
[7] 刘柏年, 皇群博, 张卫民, 等. 集合背景误差方差中小波阈值去噪方法研究及试验[J]. 物理学报, 2017, 66(2): 85-93
[8] KOPSINIS Y, MCLANGLIN S. Development of EMD-based denoising methods inspired by wavelet thresholding[J]. IEEE Transactions on Signal Processing, 2009, 57(4): 1351-1362
[9] 林忠海, 陈宏, 李晓林, 等. 宽量程光电传感器I-U特性测试系统设计[J]. 中国测试, 2020, 46(12): 92-98
[10] 杜必强, 孙立江. 基于EMD和小波熵阈值算法的超声回波信号降噪[J]. 中国测试, 2017, 43(1): 101-105
[11] CUI X C, WEI J L, QIU X B, et al. Real-time pre-processing of the pulsed eddy current signal from continuous casting slabs[J]. Insight, 2013, 55(3): 13-141
[12] HUANG N E, SHEN Z, LONG S R, et al. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis[J]. Proceedings of the Royal Society of London Serie A-Mathematical Physical and Engineering Sciences, 1998(454): 903-995
[13] 崔伟成, 许爱强, 李伟, 等. 基于拟合误差最小化原则的奇异值分解降噪有效秩阶次确定方法[J]. 振动与冲击, 2017, 36(3): 132-137
[14] 钱征文, 程礼, 李应红. 利用奇异值分解的信号降噪方法[J]. 振动、测试与诊断, 2011, 31(4): 459-463
[15] 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.