论文部分内容阅读
地震子波的最小相位（MP）假设为不适定的反褶积问题的求解提供了前提，但是这个假设的正确性一直是值得怀疑的。长期以来，有关的研究多局限于利用地震记录的二阶统计信息即记录的自相关函授或功率谱来研究此问题。但由于二阶统计量并不包含相位信息，因而无法解决非最小相让（NMP）的物理问题。随着高阶累积量与多谱概念的引入，使得借助于高阶累积量或多谱乃至它们的切片（或投影）所携带的相让信息来识别NMP地震子波成为可能。为此，本文提出运用基于高阶累积量的MA参数估计技术提取非最小相位地震子波的方法。该方法充分利用了地震记录高阶累积量中的相位信息，从而避免了单纯依赖于优化技术而忽略物理实质的倾向。实际资料处理的结果表明，该方法在混合相位、非高斯信号的处理中具有明显的优势。
The minimum phase (MP) of a seismic wavelet is assumed to provide the premise for solving the ill-posed deconvolution problem, but the validity of this assumption has always been questionable. For a long time, many researches have been limited to the use of second-order statistics of seismic records, that is, recorded autocorrelation correspondences or power spectra to study this problem. However, since the second-order statistics do not include phase information, it is not possible to solve the physical problem of Non-Minimum Consensus (NMP). With the introduction of high-order cumulant and multi-spectrum concepts, it is possible to identify NMP seismic wavelets with the help of higher-order cumulants or multi-spectra and even the entitlement information carried by their slices (or projections). Therefore, this paper proposes a method of extracting non-minimum phase seismic wavelet using MA parameter estimation technique based on high-order cumulants. This method takes full advantage of the phase information in high-order cumulants of seismic records and thus avoids the tendency of ignoring the physical substance based solely on optimization techniques. The results of actual data processing show that this method has obvious advantages in the processing of mixed phase and non-Gaussian signal.