【摘要】 在储层四性特征及其四性关系研究的基础上 ,应用 BP神经网络方法 ,对梁家楼油田沙三中储层的物性参数 (孔隙度、渗透率 )进行了预测 ,并对其预测精度进行了检验。将神经网络解释结果与常规数理统计方法精度对比可见 ,神经网络法的参数预测精度有较大的提高 ,显示出 BP神经网络法在储层参数预测中的优势与应用潜能。更多还原

【Abstract】 On the basis of investigating reservoir’s characteristics of lithologic, physical, electrical and oil-bearing properties and the relationship of them, the prediction of the reservoir lithologic parameter (shale content) and physical parameters (porosity and permeability) were carried out with the method of BP neural network. The prediction accuracy is tested. Through the accuracy correlation between the neural network interpretation results and the conventional mathematical statistics method, it is proved that the parameter prediction accuracy of the former is greatly improved. Moreover, the advantage and application potential in reservoir parameter prediction of BP neural network is further verified.更多还原