摘要
小波分析作为一种时频分析工具,特别适用于非平稳信号的分析,并且具有多分辨分析的特点。阐述了小波阈值去噪的基本原理,对脉冲涡流这种典型的非平稳信号进行了小波阈值法去噪的仿真实验,并与传统的FIR和IIR数字低通滤波方法进行了比较,然后对实际的脉冲涡流信号进行了消噪。仿真实验结果和实际消噪结果都表明,该方法能够有效地去除脉冲涡流信号中的噪声,同时又能较好地保留原信号的特征,确保了对缺陷的精确定量,是一种对脉冲涡流信号即有效又适用的去噪方法。
Wavelet analysis is a tool of time-frequency analysis,which is especially suitable for the analysis of non-stable signal,it has the character of multi-resolution analysis.The principle of wavelet threshold de-noising is introduced.Pulsed eddy current signal,as a typical non-stable signal,is simulated with wavelet threshold de-noising.It is compared with traditional methods of IIR and FIR,then actual signal is de-noised.Simulation and actual de-noising results demonstrate that noise in pulsed eddy current signal can be removed by wavelet threshold de-noising method effectively,and the character of original signal can be retained simultaneously,which ensure defects be quantified accurately.The de-noising method is effective and appropriate for pulsed eddy current signal.
出处
《计算机工程与设计》
CSCD
北大核心
2010年第16期3715-3719,共5页
Computer Engineering and Design
基金
江西省自然科学基金项目(2007GZW1530)
航空科学基金项目(BA200908097)
关键词
脉冲涡流
信号处理
小波分析
阈值
去噪
低通滤波
pulsed eddy current
signal processing
wavelet analysis
threshold
de-noising
low-pass filtering