摘要
针对预测时如何判断建模历史数据的稳定性和新的信息如何及时补充修正原预测模型,不断提高预测精度等问题,运用灰色系统理论,建立灰色等维新息GM(1,1)模型群,准确识别系统的稳定阶段,选择合理的建模数据,建立可靠的灰色模型,提高预测值的可信度;建立等维灰数递补GM(1,1)模型群预测产油量,及时补充新息数据,约束灰平面的扩展,提高模型的预测精度。将这种方法应用于中国东部某油田年产油量的建模和预测。研究结果表明,通过考察等维新息GM(1,1)模型群的后验差比和小误差概率的变化,可以判断系统的稳定和非稳定阶段,选择合理稳定的建模数据;采用等维灰数递补GM(1,1)模型群预测产油量,提高了预测灰区间的白化度,随着灰参数的不断修正,模型预测精度不断提高。
To judge the stability of history data for modeling and improve the precision of the emendatory prediction model by new information,the grey equidimensional and new-info GM(1,1)model groups was built using grey system theory,which can accurately recognize the stable stage of the system,select reasonable data for modeling,build reliable grey model,and increase the reliability of the prediction.The application of the equidimensional grey number progressive complement GM(1,1)model groups on predicting oil production can constrain the extension of grey plane and enhance the precision.This method has been applied to the modeling and prediction of the oil production for some east oilfields in our country.The stable stage of the system can be judged,and the reasonable data for modeling can be selected by investigating the error inspection of post-sample and the small error probability.The application of the model groups on the prediction of oil production can increase the whiting value of interval grey number,and will enhance the precision of the prediction model with the modification of the grey number.
出处
《油气地质与采收率》
CAS
CSCD
北大核心
2008年第5期82-85,共4页
Petroleum Geology and Recovery Efficiency
关键词
GM(1
1)模型群
灰色预测
产油量预测
等维新息GM(1
1)模型
等维灰数递补GM(1
1)模型
GM(1,1)model groups
grey prediction
the prediction of oil production
equidimensional and new-info GM(1,1)model
equidimensional grey number progressive complement GM(1,1)model
