摘要
利用逐步回归分析、神经网络、相关滤波、协克里金和非参数回归分析等方法 ,实现了由地震资料与测井资料联合应用对孔隙度参数的平面分布预测。通过实例分析比较了各自的地质效果 。
The plane distribution of porosity has been predicted through the utilizing of seismic data and well data. Five predict methods have been used like step by step regression analysis, neural network, correlation filter, cokriging and nonparametric regression analysis. The geological effects of these methods have been compared through the actual examples, and then the different characteristics and the conditions of these methods have been summarized.
出处
《地质与勘探》
CAS
CSCD
北大核心
2001年第5期56-60,共5页
Geology and Exploration
关键词
储层参数预测
多元逐步回归分析
神经网络
相关滤波
协克里金
非参数回归分析
油藏
平面分布
reservoir parameter prediction, regression analysis step by step, neural network, correlation filter, cokriging, nonparametric regression analysis
