基于小波分析的粗晶材料超声无损评价信号处理技术

Wavelet Analysis Based Signal Processing Technique in Ultrasonic Nondestructive Evaluation of Coarse Grain Material

噪声是降低超声检测可靠性的因素之一。本文利用小波分析对粗晶材料超声检测中由晶界散射引起的结构噪声进行了分析处理。和小波变换相比， 小波包变换频域分辨率高， 在信号的分解、降噪和重构时性能比前者好。对奥氏体不锈钢中人工缺陷的Ａ 型超声信号的处理结果表明， 结构噪声被有效地滤除， 且不损失缺陷信号， 信噪比明显提高。小波分析技术用于超声等无损评价信号处理，必将大大提高检测的可靠性的灵敏度。

Noise is one of the main factors that cut down the reliability of ultrasonic testing In this work wavelet analysis is employed to analyze and process the structure noise due to grain boundaries scattering in ultrasonic testing of coarse grained materials compared with wavelet transform,wavelet packet transform has a higher high frequency resolution,leading to a more content performance in signal decomposition,de noising and recon struction Applying wavelet analysis technique to process A scan signal of ultrasonic testing of artificial defects in austenite stainless steel concludes that structure noise is filtered out efficiently with no loss of defects information,leading apparent enhancement of signal to noise ratio Nondestructive Evaluation will see more reliability and sensitivity by use of wavelet analysis technique in signal processing in the near future