论文部分内容阅读
本文从分形理论出发，对地震反射系数序列进行了分形研究，得出反射系数的功率谱并不是白噪的，在多数情况下与频率的幂指数成正比关系，即具有分形特征。作者用反射系数序列具有分形特征取代反褶积中对反射系数功率谱为白噪的假设，提出了分形反褶积，并作了详细的方法原理论证。文中列举了理论和实际资料处理实例，其结果表明，分形反褶积能较好地改善反褶积的质量，提高地震记录的分辨率和信噪比。
Based on the fractal theory, fractal theory is used to study the seismic reflection coefficient sequence. It is concluded that the power spectrum of the reflection coefficient is not white noise, and in most cases is directly proportional to the power index of the frequency, that is, it has the fractal characteristic. The authors proposed a fractal deconvolution based on the assumption that the reflection coefficient sequence has a fractal feature instead of the white noise of the reflection coefficient power spectrum in the deconvolution. A detailed methodological proof is given. The paper gives examples of theoretical and practical data processing. The results show that fractal deconvolution can improve the quality of deconvolution and improve the resolution and signal-to-noise ratio of seismic records.