【摘要】 利用叠前地震数据的振幅及旅行时信息进行叠前地震记录的全波场反演,可以得到比常规叠后波阻抗反演更丰富的储层物性参数。同其它地球物理反问题类似,叠前地震全波场反演也需要某种正演算法来合成叠前记录。因为层状弹性半空间的地震波传播规律足以模拟通常的AVO分析,这样就可用慢度法作为叠前地震记录的正演模拟方法。慢度法能够提供地震波场的完全解,其结果包含了一次波、多次波和转换波。叠前反演问题由遗传算法实现,并根据叠前反演这一特定的多参数非线性问题,设计了遗传算法的所有算子,使其能够最大限度地搜索解空间并有效地搜索到最优解。叠前反演得到的纵波速度、横波速度和密度可用于指导岩性和流体的识别。更多还原

【Abstract】 With prestack seismic amplitude and travel-timeinformation, full wave field inversion can extractmuch more physical properties of reservoir than Con-ventional post-stack impedance inversioir. Similar toother geophysical inverse problems, prestack wavefield inversion also needs a forward modeling algo-rithm to synthesize prestack seismogram. Since thewave propagation in stratified elastic half space is e-nough for AVO analysis, the slowness approach canbe used as forward modeling algorithm. It can pro-vide a complete solution of seismic wave field, inclu-ding primary reflections, multiples and conversionwaves. In order to effectively search the solutionspace and find the optimum model, the GA (geneticalgorithm) operators are defined based on the specificmulti-parameter nonlinear optimization problem. TheP and S-wave velocities and densities, which are de-rived from the waveform inversion, are used as aguide to identify lithology and fluid content.更多还原