It＇ s a problem to be solved how to de-noise the signal of blast shock wave overpressure. In the conventional methods, the high frequency of the signal is cut directly by some mathematics algorithms, such as Fourier ...It＇ s a problem to be solved how to de-noise the signal of blast shock wave overpressure. In the conventional methods, the high frequency of the signal is cut directly by some mathematics algorithms, such as Fourier Transform, but some of the useful signal will be cut together. We adopt a new method for the signal de-noising of shock wave overpressure by wavelet analysis, There are four steps in this method. Firstly, the original signal is de-compoed. Then the time-frequency features of the signal and noise are analyzed. Thirdly, the noise is separated from the signal by only cutting its frequency while the useful signal frequency is reserved as much as possible. Lastly, the useful signal with least loss of information is recovered by reconstruction process. To verify this method, a blast shock wave signal is de-noised with FFF to make a comparison. The results show that the signal de-noised by wavelet analysis approximates the ideal signal well.展开更多
文摘冻结立井爆破过程中,近区监测信号中含有的基线漂零及噪声成分对其局部特征精细化提取影响显著。在对近区井壁振动信号有效采集基础上,通过互补总体经验模态分解(complementary ensemble empirical mode decomposition,CEEMD)方法、稀疏化基线估计消噪(baseline estimation and de-noising with sparsity,BEADS)方法和隐马尔可夫模型消噪(hidden Markov model de-noising,HMMD)方法等,解决了信号中基线漂移和随机噪声消除难题,并采用交叉小波变换对校正和消噪效果进行了相关性评价。实例分析结果表明:信号中缓变的基线成分遍历信号各个模态分量的整个过程,且主要集中于低频分量中,而噪声则集中在高频分量。组合分析方法对低频基线漂零和高频噪声的处理效果好,是一种高效且相对保幅的信号分析方法,可用于批量信号数据的预处理过程。
文摘It＇ s a problem to be solved how to de-noise the signal of blast shock wave overpressure. In the conventional methods, the high frequency of the signal is cut directly by some mathematics algorithms, such as Fourier Transform, but some of the useful signal will be cut together. We adopt a new method for the signal de-noising of shock wave overpressure by wavelet analysis, There are four steps in this method. Firstly, the original signal is de-compoed. Then the time-frequency features of the signal and noise are analyzed. Thirdly, the noise is separated from the signal by only cutting its frequency while the useful signal frequency is reserved as much as possible. Lastly, the useful signal with least loss of information is recovered by reconstruction process. To verify this method, a blast shock wave signal is de-noised with FFF to make a comparison. The results show that the signal de-noised by wavelet analysis approximates the ideal signal well.
