【摘要】 利用基于岩石物理模型的最优振幅随偏移距变化(AVO)线性近似公式,构建满足近临界角条件精度要求的反射系数近似公式,在地震记录的褶积理论假设前提下,结合贝叶斯理论的最大后验概率估计方法推导了最优系数的反演目标函数,考虑到拉梅参数λ和μ对地下储层信息的敏感特性,进一步将最优系数转换为拉梅参数,实现以岩石物理模型为软约束的拉梅参数提取。通过模型试算验证了方法的有效性和提取参数的高抗噪性,最后将此方法应用于实际叠前地震道集,取得了较好的储层描述结果。该方法鲁棒性较强,可以为后续的储层预测与烃类检测提供较为可靠的数据基础。更多还原

【Abstract】 By using optimal amplitude variation with offset(AVO) linear approximation formula based on rock physics model,the reflection coefficient approximation formula which satisfies the accuracy in the condition of near critical angle incidence was constructed.On the assumption of seismic convolution theory,the inversion object function for the optimal coefficients was derived by the solution estimation method of Bayesian maximum posterior probability density.Lame parameters were sensitive to the underground reservoir,so the optimal coefficients were transferred into Lame parameters further,and the extraction of Lame parameters based on the soft constraint of rock physics model was realized.The model test verifies the validity of the method and high noise immunity of the extracted parameters.Finally,this method was applied to the real prestack seismic gather and the preferable results for reservoir description were gotten.This method is so robust that it can supply reliable data for the follow-up reservoir prediction and hydrocarbon identification.更多还原