论文部分内容阅读
利用传统的数学统计方法计算采收率具有较大的困难.在对已知油藏最终采收率分析的基础上,提出了一种基于支持向量机的油藏建模方法,将影响采收率的主要因素作为输入信息,建立诸因素与采收率之间的关系,得到采收率预测模型.利用该模型计算出的采收率与实际最终采收率基本吻合,最大拟和误差为0.142%,最大泛化误差为2.744%.仿真结果表明,该模型能较好地建立各因素与最终采收率的复杂非线性关系,反映油藏的本质,具有较高的预测精度.
It is difficult to calculate oil recovery by traditional mathematical statistics method.Based on the analysis of ultimate recovery of known oil reservoir, a reservoir modeling method based on support vector machine is proposed, Rate of the main factors as input information to establish the relationship between factors and recovery rate to obtain recovery prediction model.The recovery calculated by the model and the actual final recovery rate basically consistent with the maximum error is 0.142% and the maximum generalization error is 2.744% .The simulation results show that this model can well establish the complex nonlinear relationship between each factor and the ultimate recovery ratio, reflect the nature of the reservoir and have high prediction accuracy.