离散/网络空间中油气管网设计优化方法和体系结构研究

With the development of marginal oil and gas field and the utilization of non-conventional energy sources of coalbed methane and shale gas, the technical and economic problems of surface engineering are appearing obviously. Optimal design of oil-gas pipeline network is a complex problem of large-scale hybrid optimization and many sub problems belong to NP hard problems. The technical demand of the optimal design of oil-gas pipeline network for star-branch structure, branch-star structure and branch-branch structure are concerned in this project. Considering the constraints of number, distance and capacity, firstly multiple facility location model and multi-echelon network optimization model in discrete space are proposed and the corresponding global optimization algorithms are put forward. Then，based on the delaunany triangulation, an algorithm is established to solve the multiple facility location optimal model in network space. Meanwhile, the multi-echelon network optimization model in network space and the corresponding global algorithm are presented by analyzing the optimization strategy and the coupling relation among the variables in each stage. Thus, by revealing the relationship among the network space (tree, first/higher order delaunay triangulation or complete graph), the metric and the pipeline network structure, the mathematical principles are clarified and the solving algorithm capable to optimize multiple types of network are brought forward. By synthesizing the studies of the network optimal design in continuous, discrete and network space, a complete taxonomy of oil-gas network pipeline optimization researches and the classification scheme of mathematical models are raised by summarizing the similar characteristic of models and adopting the multi-parameters description method. Therefore, it is considered that the theoretical bases and technical supports would be provide for oil-gas network pipeline surface engineering design.

随着边际油气田开发，煤层气、页岩气等非常规能源的利用，油气田规模不断扩大，地面工程的技术性和经济性问题日益显现。油气管网优化设计是一类十分复杂的大型混合优化设计问题，许多子问题属于NP难题。本项目针对星枝状、枝星状和枝枝状油气管网最优设计需求，考虑数量、距离和容量约束条件：建立离散空间多站点布置优化和多级管网设计优化模型，提出相应的整体优化算法；基于delaunany三角网，通过网络空间多站点布置优化模型和求解算法的构建，建立网络空间多级管网设计优化模型，剖析分阶段优化策略及变量耦合关系，提出整体优化算法，揭示树、一阶/高阶三角网、完全图等网络空间形式与度量方式和管网形态的关系，阐明数学原理，形成可设计多种管网结构的优化方法。综合连续、离散和网络空间管网优化方法，归纳模型的相似特征，采用多参数描述方式，提出完备的油气管网设计优化分类学方法和模型指标体系，为油气管网地面工程设计提供理论依据。

随着边际油气田开发，煤层气、页岩气等非常规能源的利用，油气田规模不断扩大，地面工程的技术性和经济性问题日益显现。油气管网优化设计是一类十分复杂的大型混合优化设计问题，许多子问题属于NP难题。本项目针对星星状、星枝状、放射环状和枝枝状油气管网最优设计需求，考虑数量、距离和容量等约束条件，构建了布局优化模型和求解算法。首先，考虑站点布置的离散特征，在离散空间中，建立了通用的星星状管网布局优化模型，提出了整体性优化算法；建立了通用的星枝状管网布局优化模型，提出了整体性优化算法；建立了放射环状管网布局优化模型，提出了分级优化策略和算法；并对提出的离散空间中星星状、星枝状、放射环状管网优化算法进行了算例验证，证明了模型和算法的有效性。然后，基于delaunany三角网，建立了网络空间多站点布置优化模型和求解算法，获得了枝状井组划分结果；建立了网络空间枝枝状管网设计优化模型，提出优化算法；揭示了一级或二级拓扑分别采用三角网、完全图的网络空间形式与度量方式（路径度量和直线度量）和管网形态（星星状、星枝状和枝枝状）的关系；通过与实际油气田数据和分级优化结果的对比，验证了模型和算法的有效性。最后，综合连续、离散和网络空间管网优化方法，从管网形态、费用函数、状态空间、求解空间和距离度量归纳了模型的特征，采用7参数描述方式，提出了油气管网设计优化分类学方法和模型指标体系。本项目提出的管网模型和算法可以为油气管网地面工程设计提供理论依据，可用于具有类似结构的管网系统，比如陆上常规油气田管网、页岩气田管网、煤层气田管网、储气库地面管网、海上油气田、水管网、公共设施网络等实体网络，也可以用于通讯系统、物流系统等。