摘要
操作成本作为油田生产过程中的主要组成部分,直接影响着油田的开发效益。为了不断降低操作成本以提高经济收益,以特高含水水驱油田为研究背景,通过该类油田操作成本的相关影响因素建立基于多重因素的两种预测模型——线性回归预测模型和非线性误差反传(BP)神经网络预测模型,最后利用组合预测法,将这两种预测模型组合在一起,以提高预测模型的有效性和可操作性,为特高含水水驱油田控制成本提供科学合理的依据。
As the main part in production process of oilfield,operational cost directly affects the benefits of oilfield development. In order to continuously reduce operational cost and improve economic benefit,with high water-cut water flooder oilfield as background,two forecasting models were set up based on various factors affecting operational cost of such kind of oilfields,which are linear regression forecasting model and non- linear BP neural network forecasting model. The two models were combined with combined forecasting method to improve the effectiveness and operability of the model so as to provide basis for making plan on operational cost of high water- cut water flooded oilfields.
引文
[1]陈武,钟水清,唐洪俊,等.油气操作成本预测方法研究[J].钻采工艺,2006,29(5):73-76.
[2]尹康.常用统计软件关于岭回归计算原理的比较分析[J].统计研究,2013,30(2):109-112.
[3]龚安,王霞,姜焕军.基于遗传算法的BP神经网络在油田措施规划预测中的应用[J].计算机系统应用,2006(11):21-24.
[4]周超,王秀芝.组合预测在油气操作成本预测中的应用研究[J].天然气勘探与开发,2009,32(4):78-80.