基于混合遗传算法的油田注水泵站优化控制.pdf

华中科技大学
硕士学位论文
基于混合遗传算法的油田注水泵站优化控制
姓名:邵旻
申请学位级别:硕士
专业:系统分析与集成
指导教师:王乘
20060401
摘要
油田注水是我国油田开发中一种十分重要的开采方式。本文在分析了国内油田注
水系统现状的基础上,针对文南油田注水系统能耗大、水压波动严重的问题,提出了
一种基于混合遗传算法的注水泵站优化控制方案。依据该方案,从工业实际应用出发,
进行了优化算法设计和优化控制软件的开发。
本文结合注水泵性能特性和工程实际要求,以变频调速往复式水泵和多级离心式
水泵组成的注水泵站为研究对象,分析其注水效率优化问题,包括最优控制数学模型
的建立、约束条件的确定、求解算法的选择和设计、软件实现和结果分析等。泵站的
效率涉及了多种因素,是一个复杂的多元非线性函数,优化策略的研究比较困难。文
中应用最小二乘曲线法和 BP 神经网络算法拟合泵组性能特性曲线,采用实数编码的
混合遗传算法进行数值优化计算,由此建立和求解泵站最大注水效率的数学模型。
最后,作者介绍了整个优化系统软件的研制,包括软件的开发环境和主要功能模
块的设计,并应用该软件进行了实例计算和结果分析,实验结果验证了优化软件的实
用性和优化算法的正确性、可靠性。

关键词:油田注水泵站节能优化控制注水效率最小二乘曲线拟合混合遗传算法














I
Abstract
Water-flooding is an important way to oil exploitation used by many oil-fields in our
country. After analyzing the actuality of the water-flooding system in oil-fields,this paper
proposed an optimization control scheme of water injection pumping station based on
Hybrid ic Algorithm to solve the problem that the huge power consumption and
serious pressure fluctuation in Wennan Oil-field water-flooding system. According to this
scheme, the thesis designed the optimization solution algorithm and developed software on
the basis of practical application.
Combining the pump’s characteristic and the project’s real requirements, this thesis
made research of the water-flooding efficiency optimization,targeting the water injection
pumping station which posed of centrifugal pumps and frequency converter driving
reciprocating pumps, and aimed at its modeling, restricted terms establishment, solution
algorithm choice and design, software development and result analysis. The efficiency of
the pumping station is plex and multi-variable function, involving many elements
which lead to great difficulty in strategy research of efficiency optimization. In the thesis,
the author used Least Squares Curve-fitting and BP work algorithm to get the
pump’s characteristics, and chose real-coded hybrid ic algorithm