基于Kalman滤波糙率反演模型的河道洪水实时预报研究

Application of real-time roughness coefficient inverse analysis using Kalman filter theory to river flood routing

建立基于Kalman滤波理论的河道糙率反演模型。将河道不规则断面沿程概化为矩形、三角形和抛物线型断面。引入水力半径与水深的经验关系,基于圣维南方程组动量方程中惯性项与弗洛德数相关性,推求水深对糙率的偏导数,建立基于Kalman滤波理论的平原区河道糙率反演模型。选择淮河干流王家坝至鲁台子河段为例,将糙率反演模型与水动力学模型相结合进行河道洪水实时预报,并与一维水动力学模型的预报结果进行比较分析。结果表明,在预见期为6小时的实时预报中,耦合模型与原水动力学模型均取得较好的预报效果,耦合模型预报精度好于原水动力学模型,也证明了所建模型的合理性。

This paper develops a flood routing model of plain rivers with a real-time roughness coefficient updating technique based on the Kalman filter theory.With the cross sections of long river simplified into rectangle,triangle and parabola shapes,the roughness coefficient can be updated step by step in routing calculation.Based on the relationship between Froude number and inertia terms of Saint-Venant equations,a relationship between roughness coefficient and water depth was obtained.In a case study of Wangjiaba-Lutaizi reach of Huaihe river,this new model was verified by comparing its real-time forecasts with the ones by a traditional hydraulic model.Results show that the river stages forecasted six hours in advance by both models are in good agreement with the observations and that the new model shows better performances than the traditional model.