基于精确Zoeppritz方程和改进人工蜂群算法的AVO反演方法研究  

AVO inversion method based on exact Zoeppritz equations and improved artificial bee colony algorithm

作　　者：罗红梅 王长江 智龙霄 颜世翠 张志敬 LUO Hongmei;WANG Changjiang;ZHI Longxiao;YAN Shicui;ZHANG Zhijing(Exploration and Development Research Institute of Shengli Oilfield,SINOPEC,Dongying 257015,China;College of Geosciences and Engineering,North China University of Water Resources and Electric Power,Zhengzhou 450046,China)

机构地区：[1]中国石油化工股份有限公司胜利油田分公司勘探开发研究院,山东东营257015 [2]华北水利水电大学地球科学与工程学院,河南郑州450046

出　　处：《石油物探》2022年第4期659-672,共14页Geophysical Prospecting For Petroleum

基　　金：中国石油化工股份有限公司科技开发部“复杂地质体多学科协同建模与有效储层预测技术”项目(P20055-8)资助。

摘　　要：叠前AVO反演是获取储层物性参数的重要方法之一。常规基于Zoeppritz方程线性近似的AVO反演方法具有计算效率高、实用能力强的特点,但受层状介质储层参数弱小变化假设的影响,在多参数同步反演与强差异的背景介质下存在一定的应用局限性。针对以上问题,提出了基于精确Zoeppritz方程和改进人工蜂群算法的AVO反演方法。在贝叶斯理论框架下,推导了基于精确Zoeppritz方程的反演目标函数;引入了改进的人工蜂群AVO反演算法,首先借鉴粒子群算法,增加调节系数,并通过二项交叉与人工蜂群算法结合,建立全局最优解引导的人工蜂群算法,进一步地,建立适应度函数与目标函数之间的关系,将AVO反演与蜂群算法相结合,实现对纵波速度、横波速度和密度的稳定反演。理论模型数据和济阳坳陷东营凹陷中央隆起带东部某实际资料测试结果表明,该方法克服了常规线性AVO反演方法存在的局限性,具有较好的稳定性和收敛精度,反演结果与实际地质规律吻合度更高,能够进一步提取较为可靠的储层物性参数,实现储层分布特征的地震描述。Pre-stack amplitude-versus-offset(AVO)inversion is important for determining the physical parameters of reservoirs.The conventional AVO inversion method,based on approximate Zoeppritz equations,has high computation efficiency and practicability.However,the method assumes a small difference in elastic parameters between the upper and lower medium.Additionally,it has certain application limitations when used simultaneously with multi-parameter inversion in the presence of a changing background medium.To address these shortcomings,we propose a method of AVO inversion based on exact Zoeppritz equations and an improved artificial bee colony algorithm.Using Bayesian inference,the objective inversion function is derived based on exact Zoeppritz equations.The improved artificial bee colony algorithm is incorporated into AVO inversion to achieve stable P-wave velocity,S-wave velocity,and density inversion.Firstly,using the particle swarm algorithm for reference,the adjustment coefficient is introduced,and the binomial crossover is combined with the artificial bee colony algorithm to form an algorithm guided by the global optimal solution.Secondly,the relationship between the fitness function and objective function is established,and AVO inversion is combined with the bee colony algorithm.Synthetic and field data from the east central uplift belt of Dongying Sag in the Jiyang Depression are used to assess the effectiveness and feasibility of this combined method.Results show that it overcomes the limitations of the conventional,linear AVO inversion method and has good stability and convergence precision.This inversion method can be used to better estimate the physical parameters and characterize the seismic distribution characteristics of reservoirs.

关 键 词：叠前AVO反演 精确Zoeppritz方程 改进人工蜂群算法 弹性参数 储层预测 储层反演 

分 类 号：P631[天文地球—地质矿产勘探]