地震勘探中基于形态成分分析的抽油机噪声衰减方法

Well-pump noise attenuating method based on morphological component analysis

抽油机的工作运行噪声对老油区二次地震勘探有严重的影响,会大幅降低采集到的地震信号信噪比.本文在抽油机噪声干扰特征研究的基础上,利用抽油机噪声干扰和有效反射信号的空间形态特征不同,提出一种小波域基于形态成分分析的抽油机噪声迭代衰减方法.最后将所提出的方法用于合成信号及成熟油田实际采集的地震信号,以说明本文方法的有效性.

The well-pump noise has serious effects on the seismic exploration of maturing oil-field. It will reduce signal＇s signal-noise ratio. Based on the characteristics of well-pump noise, we propose a MCA-based noise attenuating method which utilize the different spatial features between well-pump noise and reflective waves. Finally, we apply proposed method to synthetic data and oil-field data to illustrate the effectiveness of proposed method.