变换的GM(1,1)模型在油田钻井成本预测中的应用

APPLICATION OF TRANSFORMED GM (1,1) MODEL TO DRILLING COST PREDICTION

通过对传统GM(1,1)模型的原始数据序列进行加速平移变换和几何平均变换,有效弱化了原始数据序列的随机性,得到了变换的GM(1,1)模型。将其应用到某钻井公司的钻井成本预测分析中,利用Matlab7.0计算软件计算得到了预测值;同时对该公司某油田的某一区块2002年～2009年的单位钻井成本进行了残差和相对误差的检验。结果表明:变换的GM(1,1)模型大大提高了预测的精度。在此基础上,对未来三年该区块的单位钻井成本进行了预测,其结果对该公司在该区块的钻井活动起到了重要的指导作用。

Through accelerating both translation and geometric average transformation for raw data sequence of traditional GM(1,1) model,the randomness of sequence is weakened,and a transformed GM(1,1) model is obtained.This model has been applied to cost prediction of a drilling company.A predictive value is calculated by Matlab 7.0 software.Meanwhile,the residuals and relative error of the unit drilling cost are tested for a certain period(2002 ～ 2009) in an oilfield of the company.Result shows that the transformed model increases greatly the prediction accuracy.Moreover,the unit drilling cost of the block in the next three years has been forecasted.