中文摘要：

电测井方法所得到的信号含有一定的随机噪声，对原始信号进行处理之前需对其进行去噪。基于经验模态分解（EMD）得到的本征模态函数（IMF）具有自适应性、完备性、可重构性以及正交性和良好的时频聚集性等特点。因此，利用IMF分量去噪对于处理信号与噪声频率混叠的情况具有很好的效果，是一种多分辨率的去噪方法。本文采用一种首先经过经验模态分解，其次通过对部分IMF分量进行软阈值去噪的方法，研究结果表明，该方法能有效去除电测井信号的噪声干扰，且效果优于传统的小波软阈值去噪结果。

英文摘要：

The signals obtained by the resistivity logging are always contaminated by random noise which should be removed before processing. The intrinsic mode functions （IMF） acquired based on the empirical mode decomposition （EMD） , have lots of advantages, such as adaptive, completeness, reconfigurability, orthogonality and good time-frequency concentration. Denoising by IMF components is effective for the signals and noises mixed in the frequency domain, and which is a multi-resolution denoising method. A two-stage method is emplyed in the paper that the signals are processed by EMD first, and then by soft- thresholding method based on wavelet. The result shows that the method could remove the noises of the resistivity logging effectively and it is better than traditional wavelet method.