基于果蝇算法的气田集输布局优化
2020-12-17王昊魏立新陈双庆
王昊 魏立新 陈双庆



摘 要: 为降低地面集输系统气田成本和提高生产效率,在站场选址、站间管线连接方式等一系列环节进行人工智能布局优化,以达到线路最短或工程建设总投资最小的目的。本文将采用一种较新的智能算法果蝇算法对气田集输系统进行布局优化。根据果蝇个体通过嗅觉寻找食物的原理,以项目总投资最小为目标函数,根据地面集输系统的要求,找到约束条件,建立优化数学模型,通过MATLAB软件实现算法的运算,并对优化结果进行仿真,找出地面布局下气田管网的最优连接方式。经过果蝇算法的优化,气田地面集输管网可以显著降低工程投资成本。通过与传统的递阶优化方法的比较,说明了该算法在收敛速度和优化精度方面的优势。
关 键 词:集输系统;布局优化;果蝇算法;对比分析
中图分类号:TE862;TE89 文献标识码: A 文章编号: 1671-0460(2020)11-2507-05
Optimization of Gas Field Gathering and Transportation
Layout Based on Drosophila Algorithm
WANG Hao, WEI Li-xin, CHEN Shuang-qing
(School of Petroleum Engineering, Northeast Petroleum University, Heilongjiang Daqing 163318, China)
Abstract: In order to reduce the cost of the gas field in the ground gathering system, and improve production efficiency, the artificial intelligence layout optimization is always carried out to achieve the shortest route or the minimum investment of the project constructionin the selection of the station site, the connection between the station and the pipeline. In this paper, the layout of gas field gathering and transportation system was optimizedby a relatively novel intelligent algorithm, drosophila algorithm. Using the principle of fruit flies searching for food through olfactory, taking the minimum total investment of the project as the objective function, the constraints were found according to the requirements of the ground gathering system, and the optimized mathematical model was established. The algorithm was run by MATLAB software and the optimization results were simulated to find the optimal gas field pipe network connection in the layout. After the optimization of the drosophila algorithm, the engineering investment costof the gas field ground gathering pipeline network was significantly reduced. Through comparison and analysis with traditional grading optimization method, the drosophila algorithm showed the advantages in terms of convergence speed and optimization accuracy.
Key words: Gathering and transportation system; Layout optimization; Drosophila algorithm; Comparative analysis
目前,氣田集输管网布局优化通常采用层次优化法[1-2],将集输系统的整体切割划分为若干部分,并依次进行优化。首先优化井组分布[3-6],其次选择管线长度最短的集气站和集气末站位置作为目标函数。层次优化的缺点是每一级优化都可能是局部最优解,但不能保证是全局最优解[7]。
本文使用的果蝇算法[8]是台湾学者潘文超在2011年提出的,它是一种启发式全局优化的进化智能算法。以果蝇的觅食行为为基础,推导出全局最优解,具有调节参数少、操作简单、性能好等优点。
本文以径向管网与分支管网相结合的气田集输系统布置为例[9-14]。……
