【摘要】 基于地质统计先验信息的随机地震反演方法是一种基于蒙特卡洛的非线性反演方法。在贝叶斯理论框架下,通过序贯高斯模拟方法(sequential Gaussian simulation,SGS)和逐渐变形算法(Gradual Deformation Method,GDM)得到基于地质统计学的先验信息,然后构建似然函数,最终利用Metropolis算法实现后验概率密度的抽样,得到反演问题的解。与确定性反演结果相比,该方法能够有效地融合测井资料中的高频信息,提高反演结果的分辨率。数值模拟试验表明:本方法的反演结果与理论模型吻合较好,具有较高的分辨率;序贯高斯模拟采用一种新的逐点模拟方式,并结合GDM,有效提高了随机反演的计算效率。更多还原

【Abstract】 Stochastic seismic inversion based on the geostatistical priori information is a Monte Carlo based strategy for non-linear inversion.It is formulated in a Bayesian framework.Firstly,the priori information can be obtained through sequential Gaussian simulation(SGS)and gradual deformation method(GDM).Then we can construct the likelihood function.Finally,we apply Metropolis sampling algorithm in order to obtain an exhaustive description of the posteriori probability density and get the inversion results.Compared with the deterministic inversion,the inversion method we proposed can effectively integrate the high-frequency information of well-logging data and have a higher resolution.According to the numerical calculations,the final results match the model well and have a high resolution.In addition,we use the sequential Gaussian simulation(SGS)in a new implementation way and combine with GDM,which can improve the calculation efficiency of inversion method effectively.更多还原