一种基于BP神经网络的调驱增油预测方法

调驱后增油效果预测是调驱措施决策和方案优化设计的重要内容.通过室内岩心物理模拟实验参数,分析了调驱增油效果及其影响因素,建立了基于物理模拟参数为学习样本的调驱效果BP神经网络预测模型.此方法对于渤海南堡35-2油田A21和B17井预测结果与实际调驱增油量统计数据间误差分别为9.89%和7.22%.由此可见,基于物理模拟参数为学习样本的调驱效果BP神经网络预测模型切实可行,预测精度较高.

Prediction method of oil increment of profile-control and flooding measures using BP neural network based on core flooding parameters

The prediction of oil production increasing effect of profile-control and flooding is important to the improvement of profile-control and flooding measures and the optimization of profile-control and flooding scheme.The oil production increasing effect and the influencing factors of profile-control and flooding are analyzed through the physical simulation experiments of cores.Furthermore,BP neural network model for the prediction of oil production increasing result of profile-control and flooding was established based on core flooding parameters.The errors between the predicted oil increment and the actual statistic oil increment of A21 and B17 wells(in Nanpu 35-2 block,Bohai Oilfield) are 9.89% and 7.22% respectively,which shows the method in this paper can effectively and accurately predict the oil production increasing effect of profile-control and flooding.