摘要
选取某油田某区块为试验区,应用灰色关联方法分析影响压裂效果的主要因素,分别采用均值变换和极差变换方法对样本库的16口井的日增油量进行预测,并通过多元回归方法建立预测模型,预测结果显示极差变换方法预测结果较为准确,预测井的绝对误差平均值为0.52 t/d,能够满足现场压裂工艺的实施要求.
Taking a block of No. 2 Oil Production Site:in Daqing oilfleld as test site, in this paper, major factors which influence fracturing effects can be analyzed through gray correlation method. Utilizing mean transformation and range transformation methods, oil increment of sixteen wells of sample data can be respectively forecasted. Prediction models can be established by multiple regression method, and the results show that predicted values of range transformation methods are more accurate. Absolute average error of predicting wells is 0.52 t,/d and the value meets the implement requirements of fracturing technique in field.
出处
《数学的实践与认识》
CSCD
北大核心
2014年第1期97-101,共5页
Mathematics in Practice and Theory
关键词
灰色关联
多元回归
产能预测
预测模型
gray correlation
multiple linear regression
productivity prediction
predictingmodels