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
|
摘要
针对强背景噪声下微弱的脉冲涡流检测信号的特征量难以准确提取的问题,提出一种基于维纳自适应滤波和主成分分析的脉冲涡流信号降噪方法。该方法首先利用自适应维纳滤波在最小均方误差意义上所具有的最优特性,对脉冲涡流信号进行预处理;再将预处理信号与参考信号进行差分以消除部分系统噪声;最后利用主成分分析提取差分信号的主成分特征,通过设定阈值选取合适数目的主成分量进行重构,得到了具有高信噪比的脉冲涡流差分信号。在Q235阶梯板试件上进行脉冲涡流检测实验,运用该方法对被噪声干扰严重的检测信号进行处理,结果表明所提方法能够有效的消除强噪声对检测信号的干扰,大幅提高信噪比,是一种有效的脉冲涡流检测信号降噪方法。
Pulsed eddy current( PEC) testing signals are rather weak and often interfered by strong background noises,which make it difficult to efficiently extract signal features. In this paper,a novel PEC signal de-noising method is proposed based on the Wiener filtering and principal component analysis( PCA). Firstly,PEC signals are preprocessed by the adaptive Wiener filtering to generate a preliminary result based on the criteria of Minimum Mean Square Error( MMSE). Then,the preprocessed signals are subtracted by a reference signal to partly eliminate the system noise. Afterwards,the principal component features of differential signals are extracted by using PCA,and adequate number of principle components at a given threshold is selected to perform the signal reconstruction scheme. Following the above de-noising procedures,a PEC signal with high signal-to-noise ratio( SNR) is finally obtained. Validation experiment is also carried out on a step wedge Q235 steel plate. The presented method is applied to process experimental PEC signals severely interfered by noises. The results show that strong noises in experimental signals are effectively eliminated and the SNR is significantly improved,proving the presented de-nosing method to be effective and reliable.
Pulsed eddy current( PEC) testing signals are rather weak and often interfered by strong background noises,which make it difficult to efficiently extract signal features. In this paper,a novel PEC signal de-noising method is proposed based on the Wiener filtering and principal component analysis( PCA). Firstly,PEC signals are preprocessed by the adaptive Wiener filtering to generate a preliminary result based on the criteria of Minimum Mean Square Error( MMSE). Then,the preprocessed signals are subtracted by a reference signal to partly eliminate the system noise. Afterwards,the principal component features of differential signals are extracted by using PCA,and adequate number of principle components at a given threshold is selected to perform the signal reconstruction scheme. Following the above de-noising procedures,a PEC signal with high signal-to-noise ratio( SNR) is finally obtained. Validation experiment is also carried out on a step wedge Q235 steel plate. The presented method is applied to process experimental PEC signals severely interfered by noises. The results show that strong noises in experimental signals are effectively eliminated and the SNR is significantly improved,proving the presented de-nosing method to be effective and reliable.
引文
[1] Sophian A,Gui Y T,Taylor D,et al. A Feature Extraction Technique Based on Principal Component Analysis for Pulsed Eddy Current NDT[J]. NDT&E International,2003,36(1):37-41.
[2] Fan M,Cao B,Sunny A I,et al. Pulsed Eddy Current Thickness Measurement Using Phase Features Immune to Liftoff Effect[J]. NDT&E International,2017,86:123-131.
[3] 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.
[4] Huang C,Wu X,Xu Z,et al. Pulsed Eddy Current Signal Processing Method for Signal Denoising in Ferromagnetic Plate Testing[J]. NDT&E International,2010,43(7):648-653.
[5] 汤宝平,董绍江,马靖华. 基于独立分量分析的EMD模态混叠消除方法研究[J]. 仪器仪表学报,2012,33(7):1477-1482.
[6] Bieber J A,Shaligram S K,Rose J H,et al. Time-Gating of Pulsed Eddy Current Signals for Defect Characterization and Discrimination in Aircraft Lap-Joints[M]//Review of Progress in Quantitative Nondestructive Evaluation. Springer,Boston,MA,1997:1915-1921.
[7] Smith R A,Hugo G R. Deep Corrosion and Crack Detection in Aging Aircraft Using Transient Eddy Current NDE[C]//Proc 5th Joint NASA/FAA/DoD Aging Aircraft Conference. 2001.
[8] He Y,Pan M,Luo F,et al. Pulsed Eddy Current Imaging and Frequency Spectrum Analysis for Hidden Defect Nondestructive Testing and Evaluation[J]. NDT&E International,2011,44(4):344-352.
[9] 赵艳明,全子一. 一种有效的小波-Wiener滤波去噪算法[J]. 北京邮电大学学报,2004,27(4):41-45.
[10] 魏广芬,唐祯安,余隽. 基于主成分分析和BP神经网络的气体识别方法研究[J]. 传感技术学报,2001,14(4):292-298.
[11] Koutsogiannis G S,Soraghan J J. Selection of Number of Principal Components for De-Noising Signals[J]. Electronics Letters,2002,38(13):664-666.
[12] Tian G Y,Sophian A,Taylor D,et al. Wavelet-Based PCA Defect Classification and Quantification for Pulsed Eddy Current NDT[J]. IEE Proceedings-Science,Measurement and Technology,2005,152(4):141-148.
[13] 段向阳,王永生,苏永生. 基于奇异值分解的信号特征提取方法研究[J]. 振动与冲击,2009,28(11):30-33.
[14] 赵学智,叶邦彦,陈统坚. 矩阵构造对奇异值分解信号处理效果的影响[J]. 华南理工大学学报(自然科学版),2008,36(9):86-93.
[15] 伍权,徐志远,肖奇. 基于LabVIEW的脉冲涡流检测实验系统[J]. 测控技术,2017,36(12):123-126,136.
[16] Xu Z,Wu X,Li J,et al. Assessment of Wall Thinning in Insulated Ferromagnetic Pipes Using the Time-To-Peak of Differential Pulsed Eddy-Current Testing Signals[J]. NDT&E International,2012,51(10):24-29.徐志远(1984-),男,通讯作者,2007年于华中科技大学获得学士学位,2012年于华中科技大学获得博士学位,现为湘潭大学机械工程学院副教授,主要研究方向为电磁无损检测,工程测试与信号处理等,xuzhiyuan@xtu.edu.cn;伍权(1991-),男,2015年于湖南文理学院获得学士学位,现为湘潭大学硕士研究生,主要研究方向为脉冲涡流检测的信号处理技术,1472261187@qq.com。-
[2] Fan M,Cao B,Sunny A I,et al. Pulsed Eddy Current Thickness Measurement Using Phase Features Immune to Liftoff Effect[J]. NDT&E International,2017,86:123-131.
[3] 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.
[4] Huang C,Wu X,Xu Z,et al. Pulsed Eddy Current Signal Processing Method for Signal Denoising in Ferromagnetic Plate Testing[J]. NDT&E International,2010,43(7):648-653.
[5] 汤宝平,董绍江,马靖华. 基于独立分量分析的EMD模态混叠消除方法研究[J]. 仪器仪表学报,2012,33(7):1477-1482.
[6] Bieber J A,Shaligram S K,Rose J H,et al. Time-Gating of Pulsed Eddy Current Signals for Defect Characterization and Discrimination in Aircraft Lap-Joints[M]//Review of Progress in Quantitative Nondestructive Evaluation. Springer,Boston,MA,1997:1915-1921.
[7] Smith R A,Hugo G R. Deep Corrosion and Crack Detection in Aging Aircraft Using Transient Eddy Current NDE[C]//Proc 5th Joint NASA/FAA/DoD Aging Aircraft Conference. 2001.
[8] He Y,Pan M,Luo F,et al. Pulsed Eddy Current Imaging and Frequency Spectrum Analysis for Hidden Defect Nondestructive Testing and Evaluation[J]. NDT&E International,2011,44(4):344-352.
[9] 赵艳明,全子一. 一种有效的小波-Wiener滤波去噪算法[J]. 北京邮电大学学报,2004,27(4):41-45.
[10] 魏广芬,唐祯安,余隽. 基于主成分分析和BP神经网络的气体识别方法研究[J]. 传感技术学报,2001,14(4):292-298.
[11] Koutsogiannis G S,Soraghan J J. Selection of Number of Principal Components for De-Noising Signals[J]. Electronics Letters,2002,38(13):664-666.
[12] Tian G Y,Sophian A,Taylor D,et al. Wavelet-Based PCA Defect Classification and Quantification for Pulsed Eddy Current NDT[J]. IEE Proceedings-Science,Measurement and Technology,2005,152(4):141-148.
[13] 段向阳,王永生,苏永生. 基于奇异值分解的信号特征提取方法研究[J]. 振动与冲击,2009,28(11):30-33.
[14] 赵学智,叶邦彦,陈统坚. 矩阵构造对奇异值分解信号处理效果的影响[J]. 华南理工大学学报(自然科学版),2008,36(9):86-93.
[15] 伍权,徐志远,肖奇. 基于LabVIEW的脉冲涡流检测实验系统[J]. 测控技术,2017,36(12):123-126,136.
[16] Xu Z,Wu X,Li J,et al. Assessment of Wall Thinning in Insulated Ferromagnetic Pipes Using the Time-To-Peak of Differential Pulsed Eddy-Current Testing Signals[J]. NDT&E International,2012,51(10):24-29.徐志远(1984-),男,通讯作者,2007年于华中科技大学获得学士学位,2012年于华中科技大学获得博士学位,现为湘潭大学机械工程学院副教授,主要研究方向为电磁无损检测,工程测试与信号处理等,xuzhiyuan@xtu.edu.cn;伍权(1991-),男,2015年于湖南文理学院获得学士学位,现为湘潭大学硕士研究生,主要研究方向为脉冲涡流检测的信号处理技术,1472261187@qq.com。-