基于Powell遗传退火精确罚函数法求解气井产量动态优化模型  

DYNAMICALLY OPTIMIZED MODEL TO SOLVE GAS WELL PRODUCTION BY USING POWELL GENETIC ANNEALING METHOD WITH EXACT PENALTY FUNCTION

作　　者：丁显峰[1] 刘志斌[1] 潘大志[1] 

机构地区：[1]西南石油大学,四川成都610500

出　　处：《大庆石油地质与开发》2010年第6期110-114,共5页Petroleum Geology & Oilfield Development in Daqing

基　　金：基金项目：国家自然科学基金项目（50874094）和四川省教育厅基金重大培育项目（07ZZ043）资助.

摘　　要：针对气井产量动态优化模型的非线性特点，提出了基于Powell遗传退火精确罚函数解法。该方法把无约束优化中的直接解法Powell方法作为一个与选择、交叉和变异平行的算子，嵌入到基本遗传算法中，在遗传算法中定义Powell算子，采用退火精确罚函数法控制罚项，得到一种求无约束优化问题全局最优解的混合遗传算法。该法避免了求解模型梯度的困难，且有效地克服了Powell方法只能搜索到局部最优解，而由人为给出多个初始点进行多次计算来求最优解时，成功概率不高的缺陷；同时又能显著提高遗传算法收敛到全局最优解的概率。通过实例应用说明该方法实用而有效，可作为气井合理生产制度调整的依据。In view of the non-linear characteristics of the dynamically optimized model for gas well production, this paper proposes Powell genetic annealing method with exact penalty function. The direct method--Powell method in unconstrained optimization is regarded as an operator paralleled with selection, crossover and mutation, and then it is embedded into the basic genetic algorithm. Powell operator is defined in the genetic algorithm, and the penalty term is controlled by using the exact annealing penalty function~ so the hybrid genetic algorithm for the global optimal solution to the unconstrained optimization problem has been obtained at last. The method avoids the difficulties of solving the model gradient and effectively overcomes only searching local optimal solution and the low success probability by artificially giving more initial point to calculate and seek the optimal solution with Powell method. At the same time, the method can significantly improve the convergence probability of the global optimal solution in the genetic algorithm. Through the case application, it is shown that the method was practical and effective; it can give the evidence of the reasonable production system for the gas wells.

关 键 词：Powell 遗传算法 精确罚函数 产量 模型 

分 类 号：TE319[石油与天然气工程—油气田开发工程]