基于神经网络的剩余油参数预测研究

发布时间：2018-05-19 05:41

本文选题：BP神经网络 + 剩余油　；参考：《东北石油大学》2017年硕士论文

【摘要】：我国油田经过一次、二次采油后,加强剩余油参数预测的研究,对提高石油采收率不仅有着可观的经济效应,而且关系到国家石油战略的安全。探索油田剩余油参数预测的高效、可靠的信息处理方法具有重要的理论意义和实用价值。在针对石油的相关研究中,剩余油研究占据着非常重要的地位。本文应用BP神经网络算法,建立有效的神经网络模型,对剩余油的孔隙度及渗透率参数预测问题进行了研究。论文首先针对油田实际数据进行数据预处理,主要包含数据归一化方法的讨论及方法的确定;然后利用BP神经网络方法,应用MATLAB软件进行编程仿真及分析。论文利用某油田实际生产数据作为输入数据,在油层深度与孔隙度、渗透率参数之间建立了非线性映射关系,以此来预测该油区范围内的油层参数。仿真实验表明,文中给出的剩余油参数预测的方法准确度较高,方法是有效的。
[Abstract]:After the primary and secondary oil recovery in China, the study on the prediction of residual oil parameters is strengthened, which not only has a considerable economic effect on improving oil recovery, but also has a bearing on the safety of the national petroleum strategy. It is of great theoretical significance and practical value to explore an efficient and reliable information processing method for predicting remaining oil parameters in oil fields. The study of residual oil occupies a very important position in the related research of petroleum. In this paper, BP neural network algorithm is used to establish an effective neural network model, and the prediction of porosity and permeability parameters of remaining oil is studied. Firstly, the paper carries on the data preprocessing to the oilfield actual data, mainly includes the discussion of the data normalization method and the determination of the method, then uses the BP neural network method, uses the MATLAB software to carry on the programming simulation and the analysis. In this paper, a nonlinear mapping relationship between reservoir depth, porosity and permeability parameters is established by using actual production data as input data to predict reservoir parameters in this area. The simulation results show that the method presented in this paper is accurate and effective.
【学位授予单位】：东北石油大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TE327;TP183

【相似文献】