摘要
介绍了一种基于测井资料来预测双孔结构碳酸盐岩储层产能的新方法。首先根据碳酸盐岩剖面中双孔结构储层的地质和测井特征,提取与储层产能密切相关的多个测井和地质参数,考虑到这些参数与产能的非线性相关关系以及产能数据的变化特点,采用BP神经网络技术建立其储层产能的预测模型,由此处理了轮南地区的多口井测井资料。所预测的储层段产能与试油产能较为一致,效果良好。
A new method to predict the production output of dual pore structure reservoir in carbonate rock profile from log data is introduced in this paper. First, some parameters to reflect reservoir production are extracted in terms of geology and log characteristics. Considering the nonlinear relationship between parameters and production capability and the varying rule of these data , BP artificial neural network is adopted to create the predicting mathematical model. And log data of many wells in the Lunnan oil field of the Tarim Basin are processed.The predicting production output that is rather accordant with the practical situation shows the result is much better.
出处
《现代地质》
CAS
CSCD
2003年第1期99-104,共6页
Geoscience
基金
中国石油天然气总公司"九五"科技攻关项目(2097070352)。
关键词
测井资料
储层
地持数
神经网络
碳酸盐岩
轮南油田
塔里木盆地
reservoirs output
dual pore reservoir
log and geology parameters
artificial neural network
caved-fractured reservoir
carbonate
the Lunnan oil field
the Tarim Basin