基于大维数据驱动的油气管网泄漏监控模糊决策方法
发布时间：2019-06-12 04:38
【摘要】：输油管网状态量多及工艺复杂,难以建立精确的管网数学模型,为了能够实时监控管网的安全运行情况,本文提出一种基于大维数据驱动的管网泄漏监控模糊决策方法.首先利用管网现有的数据信息,在不对数据进行降维处理的情况下,从信息物理系统的角度出发,将油气管网的拓扑结构、阀门开度等管道物理数据以及压力、流量等运行信息数据结合起来对复杂管网系统建立数据驱动模型.然后基于大维随机矩阵谱理论,将得到的信息物理数据协方差矩阵谱分布及圆环率作为模糊决策的条件对管网运行情况进行判断.当管网拓扑发生动态变化时,提出的方法可以有效地解决误报率高的问题.最后通过仿真及实例的分析,可以证明所提出方法的有效性.
[Abstract]:It is difficult to establish an accurate mathematical model of oil pipeline network in order to monitor the safe operation of pipeline network in real time, a fuzzy decision method based on large dimensional data driving for pipeline network leakage monitoring is proposed in this paper. Firstly, using the existing data information of the pipe network, without reducing the dimension of the data, from the point of view of the information physical system, the topological structure of the oil and gas pipeline network, valve opening and other pipeline physical data, as well as pressure, flow and other operation information data are combined to establish a data-driven model for the complex pipe network system. Then, based on the large dimensional random matrix spectrum theory, the spectrum distribution and roundness of the covariance matrix of the information physical data are used as the conditions of fuzzy decision to judge the operation of the pipe network. When the topology of pipe network changes dynamically, the proposed method can effectively solve the problem of high false alarm rate. Finally, the effectiveness of the proposed method can be proved by simulation and analysis of examples.
【作者单位】： 东北大学信息科学与工程学院;
【基金】：国家自然科学基金重大项目(61627809);国家自然科学基金重点项目(61433004);国家自然科学基金(61473069,61573094) 中央高校基本科研业务费专项基金(N160404005)资助~~
【分类号】：TE973
本文编号：2497740
[Abstract]:It is difficult to establish an accurate mathematical model of oil pipeline network in order to monitor the safe operation of pipeline network in real time, a fuzzy decision method based on large dimensional data driving for pipeline network leakage monitoring is proposed in this paper. Firstly, using the existing data information of the pipe network, without reducing the dimension of the data, from the point of view of the information physical system, the topological structure of the oil and gas pipeline network, valve opening and other pipeline physical data, as well as pressure, flow and other operation information data are combined to establish a data-driven model for the complex pipe network system. Then, based on the large dimensional random matrix spectrum theory, the spectrum distribution and roundness of the covariance matrix of the information physical data are used as the conditions of fuzzy decision to judge the operation of the pipe network. When the topology of pipe network changes dynamically, the proposed method can effectively solve the problem of high false alarm rate. Finally, the effectiveness of the proposed method can be proved by simulation and analysis of examples.
【作者单位】： 东北大学信息科学与工程学院;
【基金】：国家自然科学基金重大项目(61627809);国家自然科学基金重点项目(61433004);国家自然科学基金(61473069,61573094) 中央高校基本科研业务费专项基金(N160404005)资助~~
【分类号】：TE973
【相似文献】
相关期刊论文 前1条
1 王飞;;克拉玛依人事人才网站的设计与实现[J];新疆石油科技;2001年04期
相关硕士学位论文 前1条
1 崔海涛;仿真数据驱动的事故三维模拟技术研究[D];中国海洋大学;2014年
，本文编号：2497740
本文链接：https://www.wllwen.com/kejilunwen/shiyounenyuanlunwen/2497740.html
