用置信度加权平均方法预测油田生产动态  

Predict production performance by confidence weighted mean

作　　者：段鸿彦 高兴军[2] 程仲平[3] 杨纯东[3] 李洪海[3] 

机构地区：[1]中油西安石油勘探仪器总厂,陕西西安710065 [2]中国地质大学,北京100083 [3]中油辽河油田分公司,辽宁盘锦124010

出　　处：《特种油气藏》2003年第5期42-43,46,共3页Special Oil & Gas Reservoirs

摘　　要：根据灰色预测GM(1,1)模型基本理论,提出用待预测点之前历史上的不同时间序列作为子序列,分别对未来某一指标进行灰色预测,得出一系列的预测结果,并以灰色预测的置信度作为权重,对预测结果进行加权平均,得出最终的预测结果。应用此方法对辽河油区某水驱开发砂岩油田的年产油量进行了灰色预测,取得了比较好的效果,为油田未来若干年的开发决策提供了依据。According to basic theory of gray prediction GM(1,1) model, it has been proposed that to take the different time sequence before the point to be predicted as sub - sequence to conduct gray prediction for a future index. A series of prediction results will be obtained and weighted and averaged by using gray prediction confidence as a weight to derive the final prediction result. Annual oil production of a water - drive sandstone reservoir in Liaohe oilfield has been predicted by using this method. Good prediction result is obtained, thus providing basis for future development plan.

关 键 词：置信度 加权平均 灰色预测 油田 生产动态 

分 类 号：TE331[石油与天然气工程—油气田开发工程]