盲源反褶积方法与应用研究

The method and applied research of blind deconvolution

一般求地震反演问题的各种反褶积方法都是基于反射系数白噪和子波为最小相位的假设下提出的。由于实际中并不能完全满足这两个假设,故本文试图对非最小相位子波和不是白噪的反射系数系列进行求解。针对这种盲源反褶积问题提出了一个数学框架,并利用有效的数学方法实现了对多道地震子波和反射系数的提取。理论模拟和实际资料处理结果表明,本文的方法是可行的。

Seismic inversion problem in general seek a variety of deconvolution methods are based on the reflection coefficient of white noise and wavelet for the minimum phase assumption made.As the practice does not fully satisfy these two assumptions,Therefore, this paper attempts to non-minimum phase wavelet and not the white noise of the reflection coefficient series to solve it.In view of this blind source deconvolution problem put forward a mathematical framework,And use of effective implementation of the mathematical method of multi-channel seismic wavelet and reflection coefficient extraction.Theoretical simulation and real data processing results show that the method is feasible.