固相连接接头超声A扫描信号的小波变换

从信号处理的角度给出了小波变换的工程理解，即小波变换可以看作从原始信号中不断抽取细节，最终只留下信号主干的过程；接着用小波变换对中碳钢摩擦焊和钛合金扩散焊接头的超声Ａ扫描信号进行了降噪处理。结果表明，在对小波系数的合理解释情况下，小波变换能有效地提高信噪比，并能提供了比传统方法更为完整的时域和频域信息。

WAVELET TRANSFORM OF ULTRASONIC A-SCAN SIGNALS: TESTING SOLID-STATE WELDED JOINTS

The common perceptions of wavelet transform for engineering are given from the point of view of signal processing. Wavelet transform is a process of keeping extracting details from primitive signal until the main is left. Ultrasonic A-scan signals from friction-welded joints in medium carbon steel and diffusion-bonded joints in titanium were processed by wavelet transform to reduce noises. Results showed that wavelet transform could be used to improve signal-to-noise ratio and to provide more perfect time-domain and frequency-domain information than conventional signal processing methods.