摘要
根据毛管压力曲线和相对渗透率的分形几何公式计算孔隙的分维数;以储层测井信息所反应的储层地质特性为基础,利用模糊识别方法,建立了该区块志留系低孔渗储层测井信息的模糊隶属函数和隶属方程;所得隶属度的大小可以用来判断储层的所属类型。在储层四性特征及其关系研究的基础上,应用BP神经网络方法建立了识别储层类型的测井分类模型,并对其预测精度进行了检验。综合这几种方法对目标区块10余口井进行分析处理,取得了很好的应用效果。研究认为孔隙结构分维数能定量表征砂岩的储集性能,并能反映砂岩孔隙结构的成因特征;不同成因的孔隙结构具有不同的分维数,因此可用分维数对砂岩孔隙结构进行分类和评价。
The expression of the capillary pressure and relative permeability can be used to calculate the fractal dimension of the reservoir rock pore structure. geological characters of Reservoir are revealed by logging data. Based on this information and the method of fuzzy identification, fuzzy belonging-functions and equations are derived for the low porosity and low permeability reservoirs of Silurian in the block. The types of reservoirs can thus be determined according to the belonging degree. On the basis of investigating reservoir′s characteristics of lithologic, physical, electrical and oil-bearing properties and the relationship of them, a logging classification model of identifying reservoir type is established by use of BP neural network method, and the prediction accuracy is tested. The assessment parameters are analyzed from 10 wells in target zone with the methods, the result is much better than ever before. The research indicates fractal dimensions of the pore structure can quantitatively characterize the reservoir property of sandstones. The types of the pore structure with different origin have different fractal dimensions, so fractal dimension is of importance in classification and evaluation of the pore structure in sandstones.
出处
《西南石油大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2007年第5期8-12,共5页
Journal of Southwest Petroleum University(Science & Technology Edition)
基金
国家重点基础研究发展计划973项目(2006CB202300)
关键词
分维
孔隙结构
毛管压力
相对渗透率
模糊识别
BP神经网络
fractal dimension
pore structure
capillary pressure
relative permeability
fuzzy identification
BP neural network
