基于高斯先验和马尔科夫随机场约束的非线性叠前地震反演研究及应用  

Research and application of nonlinear pre-stack seismic inversion based on Gaussian prior and Markov random field constraints

作　　者：肖爽 巴晶[1] 符力耘 郭强 张琳[1] 雒聪 XIAO Shuang;BA Jing;FU Li-yun;GUO Qiang;ZHANG Lin;LUO Cong(Hohai University,Nanjing 211100,China;China University of Petroleum(East China),Qingdao 266580,China;China Jiliang University,Hangzhou 310018,China)

机构地区：[1]河海大学,南京211100 [2]中国石油大学(华东),青岛266580 [3]中国计量大学,杭州310018

出　　处：《地球物理学进展》2020年第6期2250-2258,共9页Progress in Geophysics

基　　金：国家自然科学基金面上项目(41974123);江苏省双创计划;江苏省杰出青年基金(BK20200021);国家科技重大专项(2017ZX05008-007)联合资助。

摘　　要：叠前地震反演可以同步反演出纵横波速度和密度等弹性参数信息,在岩性分类和流体识别领域发挥重要作用.但是叠前多参数同步反演是一个不适定问题,难以获得较精确的结果.在反演优化算法方面,线性优化算法易于陷入局部极值,且强烈依赖初始模型,非线性优化算法可以避免这些问题.常规模拟退火算法在多参数同步反演过程中稳定性较差,难以获得理想的多参数结果.本文基于纵向高斯先验约束和横向马尔科夫随机场约束建立反演目标函数,能有效降低反演的不适定性和噪声的干扰.此外,提出基于多变量高斯分布的快速模拟退火算法,提高多参数同步反演的稳定性.最后,通过合成数据和工区数据测试,结果显示三参数反演剖面具有较好的横向连续性,反演结果与测井数据较吻合,表明本反演方法具有较高的分辨率和稳定性,具有一定的实际应用价值.Pre-stack seismic inversion can simultaneously invert the multiple elastic parameters, such as P-wave and S-wave velocities and density, and has been playing an important role in lithology classification and fluid identification. However, pre-stack multiparameter simultaneous inversion is an ill-posed problem and it is difficult to obtain accurate results. In terms of inversion optimization algorithm, linear optimization algorithm suffers from being trapped into local extreme and it strongly depends on the initial model. Non-linear optimization algorithm can avoid these problems. In particular, conventional simulated annealing algorithm has poor stability in the procedure of multiparameter simultaneous inversion, so it is difficult to obtain stabilized multiparameter results. In this paper, the objective function of prestack inversion is established based on the Gaussian prior constraint and the transverse Markov random field constraint in the lateral and vertical directions, respectively, which can effectively mitigate the ill-posedness and noise interference. In addition, the fast simulated annealing algorithm based on multivariate Gaussian distribution is proposed to improve the stability of multi-parameter simultaneous inversion. Finally, through synthetic data and work area data tests, the results show that the three-parameter inversion profile has good lateral continuity and the inversion results are consistent with the well logging data, indicating that the proposed inversion method has high resolution and good stability, and can be applied in practice.

关 键 词：高斯先验约束 马尔科夫随机场约束 模拟退火 叠前三参数同步反演 

分 类 号：P631[天文地球—地质矿产勘探]