论文部分内容阅读
本文在单河道卡尔曼滤波实时交替校正研究成果的基础上,将河网分解成流量边界型、水位边界型、内河道型及节点型4类基本单元,并按基本单元类型将单河道交替实时校正模式扩展到河网水动力模型中,实现了河网水流与状态变量实时校正的联动计算。与河网信息资料类型及其系列长短各异的特点相匹配,模型中变量校正模式选择灵活,对每一校正模式依据基本单元类型自动切换到相应的处理模块,程序为每一校正时刻、每一校正模块动态地创建相应维数的卡尔曼滤波器。太湖流域平原河网实效分析表明,基于环状河网的求解机制,水位校正量通过节点逐渐扩散到局部河网,实现了对河网系统的校正,并着重分析了水位校正对局部河网水量不平衡、水流运动状态的影响,证明了在平原河网洪水预报中实施变量交替滤波校正的可行性。
Based on the results of single-channel Kalman filter real-time alternating calibration, the paper decomposes the river network into four types of basic units: flow boundary type, water boundary type, inland river type and node type. According to the elementary unit type, The calibration model is extended to the hydrodynamic model of the river network, which realizes the linkage calculation of the river network current and the real-time correction of the state variables. And river network information types and their characteristics of different series match the length of the model variable correction mode selection flexibility for each correction mode based on the basic unit types automatically switch to the appropriate processing module for each correction time, A correction module dynamically creates a Kalman filter of the corresponding dimension. The actual river network analysis of Taihu Lake Basin shows that based on the solution mechanism of annular river network, the water level correction amount gradually spreads to the local river network through the nodes, which realizes the correction of the river network system. The correction of the influence of water level correction on the local river network Imbalance and water flow motion, and proves the feasibility of alternating variable filter correction in plain river network flood forecasting.