摘要
注水管网运行多年后,管网内部摩阻系数与原始设计摩阻系数有较大差别,导致计算结果与实际结果不一致。由于在生产中不能对其直接测量,通过模糊聚类方法建立了部分压力未知情况下油田注水管网模型,针对油田注水管网节点压力部分未知情况下管元摩阻系数的反演求解,分别建立了注水管网节点压力仿真数学模型和管元摩阻系数反演的最优化数学模型。通过拟牛顿法对节点压力仿真和粒子群算法对管元摩阻系数反演相结合,对节点压力部分未知的理想管网管元摩阻系数反演,实现了管元摩阻系数部分精确求解和部分预估求解。实例计算结果表明,管网水力计算结果与测试结果误差较小。研究结果可在生产实践中为管网的进一步分析提供指导。
After many years of operation of the water injection pipe network,the internal friction coefficient of the pipe network is quite different from the original one,resulting in inconsistent calculation results compared with actual results.Because the friction coefficient cannot be directly measured in production,a fuzzy clustering method is used to establish an oilfield water injection pipe network model when partial pressures are unknown.For the inverse solution of the pipe element friction coefficient when the oilfield water injection pipe network node pressure is partially unknown,the mathematical model for node pressure of water injection pipe network and the optimization mathematical model of inversion of pipe element friction coefficient are established respectively.Combining the simulation of node pressure with quasi-Newton method and the inversion of pipe element friction coefficient by particle swarm algorithm,the inversion of the pipe element friction coefficient of the ideal pipe network with partially unknown node pressures achieves partial accurate calculation and partial estimated solution of the pipe element friction coefficient.The calculation result of the case study shows that the error between the hydraulic calculation result of the pipe network and the test result is small.The study results can provide guidance for further analyses of the pipe network.
作者
任永良
孙凯
高胜
Sam Reifsnyder
姜民政
张国鹏
王妍
Ren Yongliang;Sun Kai;Gao Sheng;Sam Reifsnyder;Jiang Minzheng;Zhang Guopeng;Wang Yan(Machinery Science and Engineering College,Northeast Petroleum University;Department of Civil and Environmental Engineering of California Irvine University)
出处
《石油机械》
北大核心
2020年第8期97-103,122,共8页
China Petroleum Machinery
基金
国家重点研发计划项目“能源与水纽带关系及高效绿色利用关键技术”(2018YFE0196000)
东北石油大学引导性创新基金项目“基于物联网的油田注水系统生产运营智能调控技术”(15071120415)。
关键词
粒子群算法
注水管网
摩阻系数
拟牛顿法
反演
particle swarm optimization algorithm
water injection pipe network
friction coefficient
quasi-Newton method
inversion
