Mobile version of Hanspub

文章引用说明 更多>> (返回到该文章)

许立全, 李秀生. 不同产能评价方法的对比分析[J]. 油气井测试, 2005(6): 15-17, 72-73.


  • 标题: 基于BP神经网络产能预测方法在环江地区长6段的应用Implement of Reservoir Productivity Prediction Based on BP Neural Network in Huanjiang Chang 6

    作者: 喻思羽, 陈小东, 丁黎, 张瀚丹, 谯嘉翼

    关键字: 神经网络, 环江地区, 长6, 储层产能预测Neural Network, Huanjiang Area, Chang 6, Reservoir Productivity Prediction

    期刊名称: 《Advances in Geosciences》, Vol.6 No.2, 2016-04-22

    摘要: BP神经网络是一种基于误差反向传播的多层前馈网络技术,能很好解决非线性映射和数据泛化等问题,且能容许个别样本出现错误。环江地区长6储层产能预测影响因素较多,在综合考虑各影响因素的基础上,利用灰色关联法选择影响产能的储层参数,进而建立BP神经网络预测模型对产能进行预测,通过与基于RQI储层品质因子及支持向量回归机的产能预测方法进行对比分析,表明基于BP神经网络预测结果正确率高于另外两种方法。 Back Propagation Neural Network is a multilayer feed-forward network technology based on the error back propagation, which can be a better way to solve the problem of nonlinear mapping and data generalization, and allows some errors of the sample. The prediction of Huanjiang area chang 6 reservoir productivity has many factors. On the basis of considering all the factors, grey connection method is used to choose reservoir parameters that can affect the productivity, and then the BP neural network forecasting model is established to predict productivity. Through analysis on comparing reservoir quality factor based on the RQI with the support vector machine productivity forecasting method, it shows that the forecasting result accuracy rate based on the BP neural net-work is higher than the other two methods.