计算机工程与应用 ›› 2017, Vol. 53 ›› Issue (2): 260-264.DOI: 10.3778/j.issn.1002-8331.1608-0331

• 工程与应用 • 上一篇    下一篇

基于EMD阈值算法的脉冲涡流信号消噪

杨博楠1,张智军1,肖冰松1,江良英2   

  1. 1.空军工程大学 航空航天工程学院,西安 710038
    2.五三一一工厂 质量安全部,南京 211000
  • 出版日期:2017-01-15 发布日期:2017-05-11

De-noising of pulse current eddy signal based on empirical mode decomposition threshold

YANG Bonan1, ZHANG Zhijun1, XIAO Bingsong1, JIANG Liangying2   

  1. 1.College of Aeronautics and Astronautics Engineering, Air Force Engineering University, Xi’an 710038, China
    2.Department of Quality and Safety, Five Three One One Factory, Nanjing 211000, China
  • Online:2017-01-15 Published:2017-05-11

摘要: 针对脉冲涡流信号夹杂着较多的高频噪声,提出了一种新的经验模态分解阈值消噪算法。首先将信号分解为多个本征模态函数(Intrinsic Mode Function,IMF),对信噪比低的高频IMF进行减小噪声能量处理后得到重组信号;再对重组信号进行EMD分解后根据白噪声统计特性对IMF筛选,对噪声含量多的IMF进行小波阈值消噪;最后将处理过的IMF与噪声含量少的IMF重构得到消噪后的信号。实验仿真的结果和数据表明,该方法可以减少失真,获得更高的信噪比,能够较好地消除噪声的干扰恢复出原始的信号。

关键词: 脉冲涡流信号, 经验模态分解, 小波阈值消噪

Abstract: Since pulsed eddy current signal is mixed with more high frequency noise, it proposes a new empirical mode decomposition threshold de-noising algorithm. Firstly, the signal is recombined through EMD based on energy reduction of low SNR high-frequency IMF, and then the recombined signal is decomposed by EMD, and IMFs of much noise according to statistical properties of the white noise are denoised by the wavelet, finally, denoised IMFs and noise content less IMFs are reconstructed into the de-noised signal. Experimental data and simulation results show that this method can reduce the distortion, get higher SNR, and better eliminate the interference noise to recover the original signal.

Key words:  pulse current eddy signal, empirical mode decomposition, wavelet threshold de-noising