论文部分内容阅读
以往由测井资料确定储层参数均采用简化模型所建立的测井响应方程，因而难以反映储集层岩石孔隙结构的客观特征。鉴于神经网络具有非线性处理能力及其它一些优点，故本文采用神经网络技术定量求取储层的孔隙率和渗透率参数。文中采用BP网络模型，学习样本采用8种测井数据。网络学习结果与岩心分析结果逐点对应，关系一致，71个点的平均绝对误差仅为1．09％。
In the past, well logging data was used to determine the reservoir parameters using the simplified model to establish the well logging response equation, so it is difficult to reflect the objective characteristics of the reservoir rock pore structure. In view of the neural network has nonlinear processing capacity and other advantages, this paper uses neural network technology to quantitatively calculate the porosity and permeability parameters of the reservoir. In this paper, the BP network model is used, and the learning samples use 8 kinds of well logging data. The network learning results correspond point by point with the core analysis results, and the relationship is consistent. The average absolute error of 71 points is only 1.09%.