基于模拟退火算法预测储层参数
详细信息 本馆镜像全文    |  推荐本文 | | 获取馆网全文
摘要
讨论了如何将地震资料、测井信息和先验地质知识有机地结合起来预测储层参数,以便揭示储层内部细节和纵横向变化特征。通过将模拟退火算法与人工神经网络相结合,能够解决非线性全局优化问题,勾画出地震信息和储层参数之间的复杂关系,为储层特征描述和储层横向预测提供提供直观、可靠的地层参数剖面 -孔隙度、砂泥质百分含量、含水饱和度及渗透率等剖面。
Through integrating seismic data,borehole information,and a prior geologic knowledge organically, the article discusses the prediction method of reservoir parameter in order to reveal internal details and vertical-lateral variation features of reservoir .With the fusion of Artificial Neural Networks and simulated annealing arithmetic,we can solve the question of nonlinear optimization and obtain complicated relation of seismic information and reservoir parameter. It will supply intuitionistic and reliable stratigraphic parameter profiles,i.e.porosity,sand-shale content,hydrous saturation,and permeation profiles,etc., for geologist to do reservoir characteristic description and reservoir lateral prediction .
引文
[1][美]包约翰.自适应模式识别与神经网络[M].北京:科学出版社,1990
    [2]康立山等,非数值并行算法[M].北京:科学出版社,1994
    [3]美国勘探地球物理学家学会第63届年会论文集[C],北京:石油工业出版社,1994
    [4]黄德双.神经网络模式识别系统理论[M].北京:电子工业出版社,1996
    [5]D.S.Johnson,et.al.Optimizationbysimulatedannealing:Anex perimentalevaluation[C],Part1,AT&TBellLaboratories,Mur rayHill(NJ),1987

版权所有:© 2023 中国地质图书馆 中国地质调查局地学文献中心