论文部分内容阅读
本文在H.W.Chen及G.Chavent利用最优化理论研究油气藏生产动态拟合的基础上,通过对伴随函数空间的研究,将辨识理论由单重介质系统扩展至双重介质系统,并针对原方法多解,迭代中易发生逸出、振荡、仅收敛于局部极小等困难,重点研究和探讨了改善问题的适定性和加速算法的收敛性问题。 应用改进的方法,我们对“假想油藏模型”及实际油气田进行动态拟合及参数识别,均取得较好的结果,说明方法是可信的。
Based on the research of HWChen and G.Chavent using the optimization theory to study the dynamic fitting of oil and gas production, this paper expands the identification theory from single-medium system to dual-medium system through the research on the concomitant function space. Solution, iteration easy to escape, oscillation, only convergent to the local minima and other difficulties, focusing on the study and explore the fitness to improve the problem and accelerate the convergence of the algorithm. Applying the improved method, we get better results by means of dynamic fitting and parameter identification of “hypothetical reservoir model” and actual oil and gas fields, which shows that the method is credible.