基于小波包分析木材声发射信号消噪处理

Acoustic Emission Signals Denoising of wood based on wavelet packet analysis

由于目前的滤噪技术在检测木材声发射信号时还存在一定的缺点,本文根据木材声发射信号的特征,基于小波包分析研究了利用信号的小波包分析、计算和最佳小波包基的选取。采用默认阈值方法处理小波系数,通过小波包重构得到消噪后的木材损伤声发射信号,噪声得到较好的抑制。结果表明用小波包变换进行消噪处理,噪声消除彻底,提高了损伤缺陷检测的准确性。

The conventional AE transmitting technology fails to offer effective prediction for wood defects because of present noise processing technology. In this paper, according to the characteristic of wood defects acoustic emission signals, the noise processing method of wavelet packet transform is selected, The signal analysis of wavelet packet analysis, calculation and the choice of wavelet optimal basis is studied. The denoised signals which were reconstructed by the wavelet packet filtered by the threshold value clearly showed the wood defects characteristic frequency. The conclusion is that the use of wavelet packet transform gives a better noise processing and a more precise prediction.