An application of the BP neural network to carbonate karst reservoirs prediction
Effective porosity is one of the most important parameters in reservoir predication, especially in the carbonate karst reservoirs. In contrast to the calculated results by conventional statistical models, the BP neural network model can predict the porosity of reservoir more accurately because of its high nonlinear mapping ability and very strong abilities of self-adaptation and self-study. In this article, the author unified the different sampling interval of seismic and well logging responses by the mathematical method. Then discussed the correlation of them by the multiple linear regression. On that basis, the authors established the BP neural network model to predict the effective porosity of the reservoirs. The results shows that the porosity and the developed zone of fracture can be predicted in combination of three attributes of seismic and well logging data, moreover, the result is comparatively consistent well with the actually measured porosity and the well performance in study area.
carbonate karst reservoirs seismic responses well logging responses correlation multiple linear regression BP neural network
Yixin Yu Jinchuan Zhang ZhijunJin
School of Energy Resources China University of Geosciences (Beijing) Beijing, China School of Energy Resources China University of Geosciences (Beijing) Beijing,China
国际会议
杭州
英文
513-516
2011-10-28（万方平台首次上网日期，不代表论文的发表时间）