Abstract: The automatic analysis and unique results are the goal of well testing.This paper proposed an optimal well test analysis method based on genetic algorithm.Genetic algorithm is a fast searching technique with the principle of ‘Survival of the fittest'.The fittest of each solution is evaluated to undergo the action of the crossover and mutation operators in each generation.Especially,the genetic algorithms are suited to problems with non-linearity,discontinuities,etc in each generation,all of which can cause traditional optimization schemes to fail.It is not necessary to specify initial estimates for the parameter values and objective function continuous.The method is superior to simple genetic algorithm.In order to maintain diversity in the population and sustain the convergence capacity of genetic algorithm adaptive probability of crossover and mutation are used.The optimal values of reservoir parameters can be obtained.The good well test results can be obtained by combining with automated type curve matching method of nonlinear trgress.Two field well test data demonstrated the method is efficient.

Key words: adaptive genetic, algorithm automated matching, well test analysis, reservoir parameter

摘要： 如何实现试井资料的自动分析,保证解释结果的唯一性,一直是人们在积极探索的目标.为此,提出了一种基于自适应遗传算法的试井分析最优化方法.遗传算法是一种遵循“适者生存”原则的快速搜索方法.该方法在进化过程中,评价每个解的适应度值以执行交换和变异操作;特别适合于传统优化方法难以解决的非线性和不连续等问题.该方法优越于非线性回归自动拟合试井分析方法.它不需要估计井筒和油藏参数的初值,也不要求目标函数连续可微,并优于简单的遗传算法.它是通过解群体的适应度值,自适应地改变遗传算法的交换概率和变异概率,从而保证群体的多样性和算法的收敛能力,再结合非线性回归自动拟合方法,最终获得最佳的试井分析结果.通过实测资料分析,取得了良好的效果.

关键词: 自适应, 遗传算法, 自动拟合, 试井分析, 油藏参数