基于优化理论的管网在线仿真滤波技术研究

发布时间：2018-01-07 22:19

本文关键词：基于优化理论的管网在线仿真滤波技术研究　出处：《西安石油大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着中国石油、天然气生产与应用的蓬勃发展,油气管道迅速增长,形成了大型、复杂的管道输送系统。为了保障复杂油气管道的生产、输送和供给,管网在线仿真技术得到了越来越多的应用。但是管网实时在线仿真过程中经常受到噪声的干扰,降低了在线仿真结果的精度。为了保证管网在线仿真的精确度,需要研究一种管网在线仿真滤波方法,滤除在线仿真中存在的噪声,提高管网在线仿真的准确性。本文调研了大量国内外文献,紧跟国内外非线性滤波技术研究的最新成果,并与管网在线仿真技术相结合,进行了深入研究。通过研究分析现有管网仿真方法,运用隐式法建立管网在线仿真流体动态模型,以建立的管网在线仿真模型为基础,对管网在线仿真系统中存在的噪声干扰进行深入研究。针对管网系统状态噪声以及量测噪声的干扰,以管网在线仿真系统的状态方程和量测方程为基础建立目标函数,使用优化理论将建立的滤波目标函数达到极小化,运用最小二乘估计以及极值原理对目标函数进行处理,将非线性滤波问题归结转化为非线性优化问题,进而设计出最优管道实时在线模拟器,滤除管网在线仿真系统中的干扰噪声,以保证管网在线仿真结果的精确性。利用Matlab 2010a编制“基于优化理论的管网在线仿真滤波”计算程序,并编制了扩展卡尔曼滤波计算程序,对现实中某天然气管网系统为例,进行滤波计算,将得到的滤波计算数据与无噪数据、含噪数据以及扩展卡尔曼滤波数据进行对比,证明了本文所研究的管网在线仿真滤波方法具有较好的滤波效果,且滤波结果的精度高于现在常用的扩展卡尔曼滤波方法,以此验证了本文研究的滤波方法在管网在线仿真系统的中的适用性和有效性。
[Abstract]:With the rapid development of petroleum and natural gas production and application in China, oil and gas pipelines have grown rapidly, forming a large and complex pipeline transportation system to ensure the production, transportation and supply of complex oil and gas pipelines. The on-line simulation technology of pipeline network has been applied more and more, but in the process of real-time on-line simulation of pipeline network, it is often interfered by noise, which reduces the accuracy of on-line simulation result, in order to ensure the accuracy of on-line simulation of pipe network. It is necessary to study a kind of on-line simulation filtering method to remove the noise and improve the accuracy of on-line simulation. A large number of domestic and foreign literatures are investigated in this paper. Following the latest research results of nonlinear filtering technology at home and abroad, and combining with the on-line simulation technology of pipe network, a deep research is carried out. Through the research and analysis of the existing pipe network simulation methods. The fluid dynamic model of pipe network on-line simulation is established by implicit method, which is based on the established on-line simulation model of pipe network. The noise interference existing in the on-line simulation system of pipeline network is deeply studied, and the interference of the state noise and the measurement noise of the pipe network system is discussed. The objective function is established on the basis of the state equation and measurement equation of the on-line simulation system of pipeline network, and the filter objective function is minimized by using the optimization theory. The objective function is processed by using least square estimation and extreme value principle, and the nonlinear filtering problem is converted into nonlinear optimization problem, and then the optimal pipeline real-time on-line simulator is designed. Filtering the interference noise in the on-line simulation system of pipe network. In order to ensure the accuracy of the on-line simulation results of pipeline network, the program of "online simulation filter based on optimization theory" is compiled by Matlab 2010a, and the calculation program of extended Kalman filter is compiled. For a natural gas pipeline network system as an example, the filter calculation data are compared with the noiseless data, noisy data and extended Kalman filter data. It is proved that the on-line simulation filtering method studied in this paper has better filtering effect, and the accuracy of the filtering results is higher than that of the extended Kalman filtering method. The applicability and effectiveness of the filtering method studied in this paper in the pipe network online simulation system are verified.
【学位授予单位】：西安石油大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TE973

【参考文献】