城市区域水情仿真和数据同化的理论研究与应用

发布时间：2018-01-27 14:49

本文关键词： 水文水动力耦合模型糙率反演水情数据同化先验知识广义反演卡尔曼滤波　出处：《浙江大学》2013年博士论文　论文类型：学位论文

【摘要】：目前,城市雨洪模型多数用于规划阶段的分析,若要更好地服务于城市防洪及排水系统日常管理,须建立能够实时反映排水系统真实运行状态和能对未来一段时间内水情预报的水情仿真与预报系统。本文从城市区域产汇流特性分析等方面入手,构建了一个适用于城市区域的水文水动力耦合模型,研究了模型参数反演和数据同化技术,初步搭建了一个城市水情仿真、校正与预报的实时系统。论文的主要研究成果如下： (1)提出了以矩阵化形式对水文与水动力模型进行耦合的方法,并结合多种措施对提高计算的稳定性、精度和效率进行了研究,包括：1)采用基流法和窄缝设计法进行干湿交替的处理；2)采用迭代计算处理堰闸等非线性内边界；3)赋予节点蓄水面积以加强系数矩阵的主对角占优性；4)采用迭代法结合矩阵标识法进行方程组的求解；5)研究了固定系数矩阵、只改变方程组右端项以提高计算效率的方法；6)提出了考虑连续方程和动量方程线性化处理后二阶及以上小量的计算方法。 (2)针对定床河道糙率的率定,提出了两个结合先验知识的反演模型：糙率空间分布最平滑模型和糙率估值修正最小模型。数值仿真表明：1)当糙率初值选在合理范围内时,模型受初值选取的影响较小；2)当监测信息较少时,模型亦能获得较为合理的结果,并随着监测信息的增加,反演结果趋于真解；3)模型具有较好的抗噪性,通过控制糙率空间分布平滑项或糙率估值修正项的权重,能有效抑制监测信息误差引起的数值扰动。 (3)以扩展卡尔曼滤波为基础,构造了多个河道糙率动态修正算法。数值仿真表明：结合糙率修正平滑性且以糙率和水情变量为系统状态变量的扩展卡尔曼滤波,能有效防止糙率的突变和失真。 (4)在水位、流量等水情变量的数据同化方面,提出并探索了扩展卡尔曼滤波、集合卡尔曼滤波和广义反演法三种方法。数值仿真表明：1)扩展卡尔曼滤波的数据同化效果好,可同时对糙率、水位、流量等变量进行校正；2)集合卡尔曼滤波的适用范围广,计算简便；3)广义反演法的计算原理简单,可避免状态修正过大而严重破坏原先的水量平衡关系。 (5)在上述研究成果的基础上,针对目前应用较广的城市雨洪管理模型SWMM存在的不足进行了改进,搭建了城市水情实时仿真与预报系统的结构框架,完成了核心计算程序的开发。
[Abstract]:At present, most of the urban rain flood models are used in the analysis of the planning stage, if we want to better serve the daily management of urban flood control and drainage system. It is necessary to establish a water regime simulation and forecasting system which can reflect the real operating state of drainage system and forecast the water regime in the future in real time. This paper starts with the analysis of the characteristics of production and confluence in urban areas. A hydrohydrodynamic coupling model suitable for urban area is constructed. The parameter inversion and data assimilation techniques are studied. A preliminary urban water regime simulation is built. The main research results of this paper are as follows: 1) the coupling method of hydrology and hydrodynamic model is put forward in the form of matrix, and the stability, accuracy and efficiency of calculation are studied in combination with various measures. It includes: (1) adopting basic flow method and narrow slot design method to deal with dry and wet alternately; 2) iterative calculation is used to deal with nonlinear inner boundary such as Weir and sluice; 3) giving the node water storage area to strengthen the principal diagonal dominance of the coefficient matrix; 4) the iterative method combined with the matrix marking method is used to solve the equations. 5) the fixed coefficient matrix is studied, which only changes the right end of the equations to improve the calculation efficiency. 6) A method for calculating the second order and above after linearization of the continuous equation and momentum equation is presented. (2) according to the roughness of fixed bed channel. Two inversion models combining prior knowledge are proposed: the smoothest model of roughness spatial distribution and the modified minimum model of roughness estimation. The numerical simulation shows that when the initial roughness value is selected within a reasonable range, the numerical simulation shows that the initial roughness is selected within a reasonable range. The model is less affected by the selection of initial values. 2) when the monitoring information is less, the model can get more reasonable results, and with the increase of monitoring information, the inversion results tend to true solution; 3) the model has good noise resistance. By controlling the weight of roughness spatial distribution smoothing term or roughness estimation correction term, the numerical disturbance caused by monitoring information error can be effectively suppressed. 3) based on extended Kalman filtering. The numerical simulation shows that the extended Kalman filter is based on the roughness correction smoothness and the roughness and hydrological variables are the system state variables. Can effectively prevent roughness mutation and distortion. 4) in the aspect of data assimilation of water level, discharge and other water regime variables, the extended Kalman filter is proposed and explored. The numerical simulation shows that the extended Kalman filter has good data assimilation effect, and can be corrected for roughness, water level and discharge at the same time. 2) the set Kalman filter has a wide range of applications and is easy to calculate; 3) the principle of generalized inversion method is simple, and the state correction can avoid serious damage to the original water balance relationship. 5) on the basis of the above research results, aiming at the shortcomings of the widely used urban rain flood management model (SWMM), the structure of the real-time simulation and forecast system of urban water regime is set up. Completed the development of the core computing program.
【学位授予单位】：浙江大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：TV87;TU992

【参考文献】