小波降噪在煤岩微震信号处理中的应用

Application of Wavelet Denoising in Processing the Micro-seismic Signal of Coal-Rock

根据微震信号噪声强度ε,利用FFT滤波以及缺省的阈值确定模型、Birge-Massart策略的阈值确定模型和小波包变换中的penalty阈值确定模型对原煤单轴全程压缩加载破坏实验的微震信号进行降噪处理。理论分析及降噪结果表明,FFT滤波得到的降噪信号过于平滑,抑制了很多有用的信息成分,而小波三种阈值确定模型能够更加有效去除微震信号中的噪声成分,降噪后的信号与原信号有着很好的相似性,最大限度的反应了原始信号本身的性质。

According to noise intensity of micro-seismic（MS） signal,the denoising process of coal samples＇ MS signals obtained from uniaxial compression test with full loading to destruction was carried out with the use of FFT filtering,models determined on the basis of default threshold value,the threshold value obtained based on Birge-Massart strategy and the penalty threshold value of wavelet packet transform.The theoretical analysis and results of denoising process showed that the MS signals obtained after denoising with FFT filter was so smooth that the lots of the useful information was inhibited,however,comparing with FFT filter,the models determined based on the three of wavelet threshold values were able to remove the noise better,the MS signals obtained after denoising were similar to the original MS signals and reflected the properties of the original MS signals in the maximal degree.