星状集输管网拓扑结构的整体优化

针对星状集输管网结构特点,建立以管网建设总投资为目标函数,以系统中节点连接关系、站点位置、管线参数为优化变量的星状油气集输管网拓扑结构优化模型.将蚁群算法与遗传算法相结合对模型进行整体优化求解.在蚁群算法中,将节点连接关系的确定转化为路径选择问题,将启发因子表示为管段建设成本的函数,用路径方案对应的管网建设总成本计算信息素的积累量.在遗传算法中,以格雷码形式将站址信息储存于染色体上,用蚁群算法求得每种站址分布方案下最优井组和管径,并用其计算各染色体的适应度,从而同步求得最佳站址、最佳井组划分和管线参数.结果表明,所设计算法优化质量髙于分级优化,且鲁棒性强,不受计算初始值影响....

Full description

Saved in:
Bibliographic Details
Published in中国石油大学学报（自然科学版） Vol. 40; no. 4; pp. 133 - 140
Main Author 刘刚 许继凯 国志刚 陈雷 卢兴国 滕厚兴 徐睿妤
Format Journal Article
LanguageChinese
Published 中国石油大学储运与建筑工程学院,山东青岛,266580%山东省天然气管道有限责任公司,山东济南,250101%中国石化胜利油田分公司油气集输总厂,山东东营,257000 2016
Subjects
分步优化
拓扑结构
整体优化
蚁群算法
遗传算法
集输管网
multilevel optimization
genetic algorithm
拓扑结构
遗传算法
蚁群算法
topological structure
ant colony algorithm
集输管网
分步优化
整体优化
gathering pipe network
global optimization
Online AccessGet full text
ISSN1673-5005
DOI10.3969/j.issn.1673-5005.2016.04.018

Cover

More Information
Summary:针对星状集输管网结构特点,建立以管网建设总投资为目标函数,以系统中节点连接关系、站点位置、管线参数为优化变量的星状油气集输管网拓扑结构优化模型.将蚁群算法与遗传算法相结合对模型进行整体优化求解.在蚁群算法中,将节点连接关系的确定转化为路径选择问题,将启发因子表示为管段建设成本的函数,用路径方案对应的管网建设总成本计算信息素的积累量.在遗传算法中,以格雷码形式将站址信息储存于染色体上,用蚁群算法求得每种站址分布方案下最优井组和管径,并用其计算各染色体的适应度,从而同步求得最佳站址、最佳井组划分和管线参数.结果表明,所设计算法优化质量髙于分级优化,且鲁棒性强,不受计算初始值影响.
Bibliography:The topological structure optimization model of radial patter oil and gas gathering pipe network was built accordingto its structural characteristics,with the total construction cost of the pipe network as the objective function,and the connectionrelation of nodes, pipeline parameters and location of stations as the optimization variables. To avoid the deficiency ofmultilevel optimization, ant colony algorithm and genetic algorithm were combined to solve the optimization model globally.In ant colony algorithm, the determination of connection relation was converted to the routing problem, heuristic factor wasexpressed as the function of pipe construction cost,and the total construction cost of pipe network corresponding to the routingscheme was used to calculate the pheromone accumulation. In genetic algorithm, the information of station location was storedin chromosomes using gray code, and the well-group scheme and pipe diameters were obtained by ant colony algorithm andwere used to calculate the fitness of
ISSN:1673-5005
DOI:10.3969/j.issn.1673-5005.2016.04.018