基于delta波包叠加的叠前深度偏移

发布时间：2018-10-31 18:13

【摘要】：作为高斯波束的对偶表达,delta波包在时空域具有良好的数学特性,将其引入偏移领域,可推导出新的反射地震偏移方法:基于delta波包叠加的叠前深度偏移。新方法属于射线类偏移,采用Kirchhoff绕射积分进行波场延拓,但以delta波包叠加的形式重构格林函数。因为delta波包的引入,整个成像过程可以在时间域进行,这就避免了因为高斯波束偏移在频率域可能遇见的问题;同时新方法仍具有可处理多走时和计算焦散区振幅等优点,面对复杂介质具有良好的成像能力。在成像精度上,基于delta波包叠加的叠前深度偏移可以达到高斯波束偏移的水准,但在计算速度上却有明显差距。为了提高计算效率,可以从两方面对新方法进行深度研究:一是根据delta波包的性质,对褶积运算进行近似处理。成像时,需要计算震记录(经过滤波)和delta波包的褶积,褶积运算量很大。因为delta波包满足柯西分布,所以可将波包参与的褶积运算简化为乘积运算,计算时间至少缩短20%;二是利用最速下降法推导快速算法。成像时,快速算法只需计算特定出射方向的delta波包,计算量大幅下降。这里共推导了4种快速算法:共炮域3种和共炮检距域1种;每种算法的计算速度并不相同,但过快的计算速度无法保证成像质量,所以可根据需求选择合适的快速算法。以上研究都是基于互相关成像条件,为了进行振幅保真偏移,还推导了基于delta波包叠加的反褶积成像条件。反褶积成像条件的叠加剖面与地层反射率有着更为密切的联系,在地震解释、地震反演以及速度分析中具有重要意义。
[Abstract]:As a dual expression of Gao Si beam, delta wave packet has good mathematical characteristics in space-time domain. A new method of reflection seismic migration, prestack depth migration based on delta wave packet stack, can be deduced by introducing it into migration field. The new method belongs to ray migration. The Kirchhoff diffraction integral is used for wave field continuation, but the Green's function is reconstructed in the form of delta wave packet stack. Because of the introduction of delta packet, the whole imaging process can be carried out in time domain, which avoids the problem that Gao Si beam offset may encounter in frequency domain. At the same time, the new method still has the advantages of dealing with multi-path time and calculating the amplitude of the caustic region, so it has good imaging ability in the face of complex media. In terms of imaging accuracy, the pre-stack depth migration based on delta wave packet stack can reach the level of Gao Si beam migration, but there is a significant difference in computing speed. In order to improve the computational efficiency, the new method can be studied in depth from two aspects: one is to approximate the convolution operation according to the properties of delta wave packet. It is necessary to calculate the convolution of seismic records (filtered) and delta wave packets when imaging. Because the delta wave packet satisfies the Cauchy distribution, the convolution operation in which the wave packet is involved can be simplified as a product operation, and the computing time is at least 20 shorter. The second is to deduce the fast algorithm by using the steepest descent method. In imaging, the fast algorithm only needs to calculate the delta wave packet with a specific direction of emission, and the computational complexity is greatly reduced. In this paper, four fast algorithms are derived: common shot domain (3) and common offset domain (1). The calculation speed of each algorithm is not the same, but too fast calculation speed can not guarantee the imaging quality, so we can choose a suitable fast algorithm according to the demand. All the above studies are based on cross-correlation imaging conditions. In order to carry out amplitude fidelity migration, the deconvolution imaging conditions based on delta wave packet stack are derived. The stacking profile of deconvolution imaging conditions is more closely related to the formation reflectivity, which is of great significance in seismic interpretation, seismic inversion and velocity analysis.
【学位授予单位】：吉林大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：P631.4

【相似文献】