油田地面集输系统树状管网优化算法研究

发布时间：2018-10-10 20:02

【摘要】：在油田地面建设中,集输管网的投资占了相当大的比重,因此,对集输管网的拓扑结构及运行参数进行优化设计十分必要。树状管网具有节省投资的优势,然而树状管网的拓扑形式、水力热力计算都较为复杂,且集输管网规模庞大,因此有必要对树状集输管网的优化算法进行研究。本文针对树状管网的特点,对树状集输管网优化算法进行了研究,并编制了集输系统树状管网优化设计软件。由于许多油田分布在地形起伏较大的地区,而高程变化对管网的水力、热力及管长计算具有重要影响,因此本文建立了管线三维计算模型,并在此基础上,建立了混输管线的热力及水力计算模型。本文对目前主要的油田管网集输流程进行了研究,确定了树状集输管网集输模式,对树状管网的描述方式进行了研究,建立了基于节点的树状集输管网数据结构。本文对树状管网的分级优化算法进行了研究:确定了树状集输管网分级优化策略,建立了树状集输管网分级优化模型,采用模糊聚类的方法进行最优化分组,加权平均法进行站场位置优化,分别对Prim、Kruskal及Dijkstra算法进行了理论研究并将其应用于树状集输管网的拓扑结构优化,对树状管网的分级优化算法进行了实例分析。本文对树状管网的整体优化算法进行了研究:确定了树状集输管网的整体优化策略,以费用最省为目标、可靠性为惩罚函数建立了树状管网整体优化模型。对单亲遗传算法进行了改进:最优化分组与单亲遗传算法相结合,单亲换位算子、倒位算子与移位算子相结合,单点与多点代间交叉遗传操作,并对传统的比例选择算子进行了改进。另外,本文通过生成管网初步连接图的方法提高了单亲遗传算法的遗传效果与效率,并结合管网初步连接图对单亲遗传算法的编码方式进行了改进,最终将改进后的单亲遗传算法应用于集输系统树状管网优化设计,并对单亲遗传算法的影响因素进行了分析。本文编制了集输系统树状管网优化设计软件,采用国内某油田的实际油井数据进行了实例计算,给出了不同算法的最佳优化方案,并对不同算法的优化效果及效率进行了对比分析研究。
[Abstract]:In the oil field surface construction, the investment of the gathering and transportation pipeline network accounts for a large proportion, so it is necessary to optimize the topology structure and operation parameters of the gathering and transportation network. Tree pipe network has the advantage of saving investment. However, the topological form of tree pipe network, hydraulic and thermal calculation are more complex, and the scale of gathering and transportation network is large, so it is necessary to study the optimization algorithm of tree gathering and transportation network. According to the characteristics of tree-shaped pipeline network, this paper studies the optimization algorithm of tree-shaped gathering and transportation network, and compiles the optimal design software of tree-shaped pipeline network of gathering and transportation system. Because many oilfields are distributed in areas with large topographic fluctuation, and elevation changes have an important influence on the calculation of hydraulic, thermal and pipe length of pipe networks, a three-dimensional calculation model of pipelines is established in this paper, and on the basis of this model, The thermodynamic and hydraulic calculation models of mixed pipeline are established. In this paper, the main gathering and transportation flow of oil field pipeline network is studied, the gathering and transportation mode of tree gathering and transportation network is determined, the description mode of tree pipe network is studied, and the data structure of tree gathering and transportation network based on node is established. In this paper, the hierarchical optimization algorithm of tree network is studied: the hierarchical optimization strategy of tree gathering and transportation network is determined, the hierarchical optimization model of tree gathering and transportation network is established, and the method of fuzzy clustering is used to optimize the grouping. The weighted average method is used to optimize the location of the station yard. The Prim,Kruskal and Dijkstra algorithms are studied in theory and applied to the topology optimization of the tree gathering and transportation network. The hierarchical optimization algorithm of the tree network is analyzed as an example. In this paper, the whole optimization algorithm of tree pipe network is studied: the overall optimization strategy of tree gathering and transportation network is determined, and the whole optimization model of tree pipe network is established with the aim of saving the most cost and the reliability as the penalty function. Some improvements are made to the single parent genetic algorithm: the combination of the optimal grouping and the single parent genetic algorithm, the single parent transposition operator, the inversion operator and the shift operator, the single point and multipoint intergenerational crossover genetic operation, The traditional proportional selection operator is improved. In addition, this paper improves the genetic effect and efficiency of the single parent genetic algorithm by generating the initial connection diagram of the pipe network, and improves the coding method of the single parent genetic algorithm combined with the preliminary connection diagram of the pipe network. Finally, the improved single parent genetic algorithm is applied to the optimal design of tree pipe network in gathering and transportation system, and the influencing factors of the algorithm are analyzed. In this paper, the optimal design software of tree pipe network in gathering and transportation system is developed. The actual oil well data of a certain oil field in China are used for example calculation, and the optimal optimization scheme of different algorithms is given. The optimization effect and efficiency of different algorithms are compared and analyzed.
【学位授予单位】：中国石油大学(华东)
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TE863

【相似文献】