面向断溶体缝洞储层的智能化散射波地震成像方法OA北大核心CSTPCD

To exhance the efficient exploration and development of fractured reservoirs and achieve high-precision imaging of small-scale geological bodies amidst strong reflections,we propose an intelligent method for separating and imaging reflections and diffractions.Leveraging the geometric differences between reflections and diffractions in dip-angle domain gathers,we in-itially utilize generative adversarial neural networks(GANs)to kinematically identify and separate reflections and diffrac-tions.Then,based on waveform amplitude characteristics,we adopt an adaptive subtraction method for dynamically separa-ting reflections and diffractions.Ultimately,the separated gathers are stacked to produce reflector images,depicting the sub-surface continuous impedance interfaces,and diffractor images,revealing small-scale geological bodies.Numerical experi-ments conducted on synthetic and field data validate the efficacy of the proposed method in enhancing imaging accuracy of fault karsts and achieving high-resolution seismic imaging of small-scale geological targets.