水动力学模型卡尔曼滤波实时校正技术

Real time updating of hydrodynamic model by using Kalman filter

本文采用卡尔曼滤波技术实现圣维南方程实时校正以提高水动力学模型应用于洪水预报的精度水平。首先利用有限差分方法将圣维南方程进行离散化并线性化,选择不同的状态变量组建了三种卡尔曼滤波状态空间方程,对预报进行实时校正。选择长江干流清溪场至万县河段进行试算。

The Kalman filter was applied to realize the real time updating of Saint-Venant equation for promoting the accuracy of flood forecasting. The discretization and linearization of Saint-Venant equation were carried out by means of finite difference method. Three kinds of Kalman filter spatial equation were established by using different state variables for real time updating. The calculation of flood forecasting for river section from Qingxichang to Wanxian of Yangtze River verifies that the proposed method can remarkably promote the accuracy of forecasting.