基于成本动因的油田区块成本预测方法研究

发布时间：2018-06-10 01:33

本文选题：油田区块 + 成本动因　；参考：《中国石油大学(华东)》2014年硕士论文

【摘要】：随着国内油田开发的不断深入,许多油田区块进入开发的中后期,含水率和开采难度不断上升,为维持产量水平需要投入巨额成本来弥补自然递减,油田区块成本不断攀升。目前区块成本预测的主要做法是参照历史水平并根据本年的计划来进行预测,这种做法缺乏可靠性和准确性。因此,寻找适合油田实际情况并且预测精度较高的成本预测方法,对于促进区块成本的控制和经济效益的提高具有重大意义。基于以上背景,本文对油田区块成本预测方法进行探讨。首先对油田区块成本的构成及特性进行了分析,在此基础之上,对成本动因进行了选择,得出了基于成本动因的成本函数,并建立了成本动因合并模型,为进行成本预测奠定了基础。然后,结合油田区块成本预测的要求和对成本预测方法的评价分析,从诸多方法中选择了BP神经网络预测方法。最后,以油田区块油气提升系统成本预测为例,建立BP神经网络预测模型进行预测,并就预测结果与回归预测法、指数平滑法、移动平均法进行了对比分析,得出BP神经网络预测有利于提高油田区块成本的预测精确性。在具体的应用中,还应该加强成本动因数据的基础工作,完善信息管理系统,进一步提高油田区块成本预测水平。
[Abstract]:With the deepening of domestic oilfield development, many oil field blocks enter the middle and late stage of development, the water cut and the difficulty of exploitation are rising constantly. In order to maintain the production level, it is necessary to invest a huge amount of cost to make up for the natural decline, and the block cost of the oil field is constantly rising. At present, the main method of block cost prediction is to forecast according to historical level and according to this year's plan, which lacks reliability and accuracy. Therefore, it is of great significance to find a cost forecasting method suitable for the actual situation of oil field and to improve the economic benefit and control of block cost. Based on the above background, this paper discusses the prediction method of oil field block cost. Firstly, the composition and characteristics of oil field block cost are analyzed, then the cost driver is selected, the cost function based on cost driver is obtained, and the cost driver combination model is established. It lays a foundation for cost prediction. Then, according to the requirements of oil field block cost prediction and the evaluation and analysis of cost forecasting methods, BP neural network forecasting method is selected from many methods. Finally, taking the cost prediction of oil and gas lifting system in oilfield block as an example, a BP neural network forecasting model is established, and the prediction results are compared with regression prediction method, exponential smoothing method and moving average method. It is concluded that BP neural network prediction is helpful to improve the accuracy of oil field block cost prediction. In the specific application, we should strengthen the basic work of cost driver data, perfect the information management system, and further improve the level of oil field block cost prediction.
【学位授予单位】：中国石油大学(华东)
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F426.22;F406.7

【参考文献】