中文摘要：

提出了一种针对小波地震信号的新型去噪方法,该方法结合了非线性小波阈值去噪算法和递归最优阈值选择算法.阐述了非线性小波阈值去噪算法的基本原理和递归最优阈值选择算法的步骤,并与传统的数据平滑法和流行的基于MAD的小波固定阈值去噪法进行了深入比较.仿真实验和统计数据表明,本方法既保证了处理后数据序列的光滑性,又保证了真实信号的完整性;克服了其他两种方法的不足,在处理效果上取得了较大优势,并在实际震相的判读工作中证明了本方法的有效性.

英文摘要：

A new denoising method of seismic signals based on nonlinear wavelet thresholding denoising and recursive optimal thresholding was proposed.The method was compared with data smoothing and wavelet fixed thresholding denoising method via MAD.Simulations and statistic evidences demonstrated that the presented approach could keep the smoothness of processed data and the integrity of real signals.It conquered the shortages of data smoothing and wavelet fixed thresholding denoising,and had much better efficiency and good application value.