电磁超声检测信号的小波自适应阈值降噪研究

Study on Wavelet Adaptive Threshold De-noising for Electromagnetic Acoustic Detection Signals

针对电磁超声检测信号中通常混有大量噪声干扰,使用小波软硬阈值降噪后信号存在震荡失真、细节丢失等问题,构造了一种新的自适应阈值函数。该自适应阈值函数处处连续且高阶可微。试验结果表明,对电磁超声检测信号进行小波自适应阈值处理,不但可以保护信号的细微特征不被消除,防止信号震荡,增强光滑性;而且还能减小均方误差和失真,提高信噪比,从而提升电磁超声检测的可靠性和准确性。

Usually, there are a large amount of noise interference existing in the detection signals of electromagnetic acoustic { EMA ） , the problems, e.g. oscillation distortion, detail loss, etc. , may exist in the signals that de-noised by wavelet soft-hard threshold. In order to solve these problems, a new self-adaptive threshold function is established. This function is continuous everywhere and also higher-order differentiable. The experimental results show that the EMA detection signals processed by wavelet self-adaptive threshold, the detail features are not removed, and the oscillation of signals is avoided, the smoothness of signals is enhanced, the means square error and distortion are decreased while the signal noise ratio is increased, thus the reliability and correctness of EMA detection are elevated.