摘要
在裂缝和溶孔发育井段由成像测井解释得到的面孔率来刻度常规测井,利用径向基函数神经网络进行训练,初步得到不同的常规测井曲线的相对权系数组合,确定出常规测井曲线对裂缝贡献的相对权系数,同时还建立了成像测井求取的裂缝孔隙度与裂缝常规测井孔隙度的经验关系,它们具有线性相关关系。
In fracture and vugular pore developed interval, the ratio of surface area to volume derived from FMI is used to calibrate conventional logging data, and the neural network is trained by basic radial function. Relative weight coefficients of different conventional logs are obtained, the relative weight coefficient of how much the conventional log contributing to fractures is determined, and an empirical relation between fracture porosities derived from FMI and from conventional logging is established. It shows that the two porosities are linear related.
出处
《特种油气藏》
CAS
CSCD
2005年第4期28-29,i0009,共3页
Special Oil & Gas Reservoirs
基金
中石化总公司科研项目"新疆塔里木盆地碳酸盐岩储层测井技术应用研究(99-111-04-01-02)"部分内容
关键词
成像测井
常规测井
裂缝孔隙度
权系数
神经网络
FMI
conventional logging
fracture porosity
weight coefficient
neural network
