油田年产量预测的数学模型  

The Mathematical Model for the Oil Field Annual Output Prediction

作　　者：何文章[1] 田婧[1] 

机构地区：[1]天津工程师范学院数理与信息科学系,天津300222

出　　处：《工程数学学报》2005年第8期72-76,共5页Chinese Journal of Engineering Mathematics

基　　金：天津市高等学校科技发展基金（20040708）.

摘　　要：本文将改进的灰色Verhulst模型用于某油田产量的预测。在平均相对误差达到最小准则或最大相对误差达到最小准则下，分别给出了估计灰色Verhulst模型中参数的线性规划方法。研究表明，模型中的背景值参数和边值修正项对模型的预测精度均有影响，在此基础上，分别以平均相对误差达到最小或最大相对误差达到最小为适应度，提出了基于遗传算法求解最佳背景值参数和最佳边值修正项的方法。最后将改进的灰色Verhulst模型用于某油田产量的实际预测，预测精度很高，结果令人满意。In this paper, an improved grey Verhulst＇s model is used to predict some oil-fields＇ annual output. Under the average relative error minimization criterion or the minimization of the maximum of relative error criterion, the linear programming method to estimate grey Verhulst model＇s parameter is derived, respectively. Considering that both the background value parameter and the boundary improvement value in the model have an effect on the prediction accuracy, we define the adapting degree as the minimization of the average relative error or the minimization of the largest relative error, and put forward a method to achieve an optimal background value parameter and an optimal boundary improvement condition based on the genetic algorithm. When used to practically predict an oil field＇s annual output, the improved grey Verhulst＇s model proves to be great with satisfactory predict accuracy.

关 键 词：灰色VERHULST模型 油田年产量预测 线性规划 遗传算法 

分 类 号：N941.5[自然科学总论—系统科学] O159[理学—数学]