【摘要】 油藏产能是衡量油藏潜在产油能力的综合指标,产能的高低直接影响油藏项目的经济效益。因此,建立有效的油藏产能预测模型,对油藏的勘探开发有重要的指导意义。油藏受到渗透率、孔隙度等若干地质因素的影响,且各个因素之间关系复杂,难以建立精确的数学表达式来描述其动态的生成过程。通过使用混合粒子群优化算法(Particle Swarm Optimization,PSO)和支持向量回归机(Support Vector Regression,SVR)的模型(PSO-SVR模型),建立地质因素与产能之间的非线性函数映射关系,以实现对油藏产能预测。通过对比实验,基于PSO-SVM的模型与实验数据的一致性,证实了其良好的性能。更多还原

【Abstract】 Reservoir productivity is a comprehensive index to measure the potential oil production capacity of a reservoir. The productivity directly affects the economic benefits of a reservoir project.Therefore,the establishment of an effective reservoir productivity prediction model has an important guiding significance the exploration and development of the reservoir. The reservoir is affected by some geological factors such as permeability,porosity and so on,the relationship between the various factors is complex. It is difficult to establish accurate mathematical expressions to describe the dynamic generation process. The purpose of this study is to establish a nonlinear function mapping relationship between geological factors and production capacity by using a hybrid PSO-SVM-based model to predict reservoir productivity. By comparative experiments,the PSO-SVM-based model is consistent with the experimental data to confirm its good performance.更多还原