摘要
在油汽探测的研究中,提高探测的准确性,结多孔介质三维重构在油气探测有着重要意义,用规方法对多孔介质中长的孔隙通道和复杂的孔隙空间形状重构效果较差。为了提高探测精度,提出一种基于多点地质统计(MPS)和多重模板的多孔介质三维重构方法。将微米精度的多孔介质体数据作为训练图像,把逐渐密集化的网格作为多重数据模板提取其空间结构特征,通过MPS将这些特征信息复制到待模拟区域得到多孔介质的拓扑结构,实现多孔介质三维重构。方法应用于砂岩样品三维信息重构。进行仿真的结果表明,方法有效地重构了砂岩中复杂的孔隙空间形状。证明了重构的多孔介质具有与实际情况的结构特征相似,有效地提高了重构的精度。
The three-dimensional reconstruction of porous media is quite significant for the exploration of oil and natural gas.However,general structure-making methods often fail to describe the long porous space and complicated porous structures.Therefore,a method is proposed using multiple-point geostatistics(MPS) and multiple-grid templates to reconstruct porous media.A 3D training image is obtained from the volume data of porous media with the resolution of microns.The gradually finer grids are taken as multiple data templates to extract characteristics of porous space,which are "copied" by MPS to the regions to be reconstructed,realizing the final reconstruction.The experiment based on the 3D reconstruction of sandstone samples proves that this method reconstructs the complicated porous structure in sandstone well.Comparison of porosity and variogram curves also demonstrates that the characteristics of porous media reconstructed by this method are similar to those of real porous media in the volume data of sandstone samples.
出处
《计算机仿真》
CSCD
北大核心
2011年第4期238-241,260,共5页
Computer Simulation
基金
浙江省自然科学基金项目(Y1080379)
浙江省教育厅科研项目(Y200803026)
关键词
多点地质统计法
多孔介质
多重模板
训练图像
条件概率分布函数
Multiple-point geostatistics
Porous media
Multiple-grid template
Training image
Conditional probability density function