基于遗传算法的机抽系统优化设计新方法  

Genetic Algorithm-based Optimizing Design for Rod Pumping System

作　　者：赵瑞东[1] 熊春明[1] 张建军[1] 吴晓东[2] 师俊峰[1] 朱拾东[1] 张娜[1] 

机构地区：[1]中国石油勘探开发研究院,北京100083 [2]中国石油大学(北京)石油工程教育部重点实验室,北京102249

出　　处：《石油天然气学报》2012年第6期158-160,170,共3页Journal of Oil and Gas Technology

基　　金：中国石油天然气股份有限公司科学研究与技术开发项目(2011B-1708)

摘　　要：机抽系统优化设计问题属于多参数系统优化问题,参数包括冲程、冲次、泵径、平均杆径、下泵深度等。将遗传算法引入到机抽系统优化设计中,形成了一种新方法,该法采用二进制编码形式来表示抽汲参数和下泵深度,按优胜劣汰的原则来寻求问题的最优解,既能提高优化设计的速度,又能克服模糊数学以及专家系统等方法易漏掉最优方案的缺点。通过实例进行敏感性分析表明,遗传代数越多、交叉系数越高时,算法优化效果越好,而适中的变异系数有利于优化设计。Genetic algorithm had very big space for optimization,powerful global search ability,high precision,strong robustness for optimization of multi-parameter system.Rod pumping optimization belonged to multi-parameter optimization,which included stroke,strokes per minute,pump diameter,average rod diameter and depth of pump.A new method was developed by introducing the genetic algorithm into the rod pumping optimization design.In the method,a binary encoding was deployed to present swabbing parameters and depth of pump for avoiding the shortcomings of omissions in fuzzy mathematics and expert system for achieving minimum system efficiency and optimal solution.The results show that genetic algorithm-based optimal designs have good optimization results and certain practical value.

关 键 词：遗传算法 有杆泵 系统效率 多参数 优化设计 

分 类 号：TE933[石油与天然气工程—石油机械设备]