Research & Application Of Flood Propagation Model

It is well known that flood may badly endanger human's lives and wealth. So effective and accurate forecasting is a very important subject, which has a pretty strong academic and practical meaning. Based on the research of flood process in Qingjiang River, flood propagation theory, mathematical model and realization of the algorithm are discussed in this paper. Also, Saint-venant equations and Muskingum model are analyzed. As a result, some useful conclusions are deduced.At first, development of flood propagation theory is reviewed, then fundamental principle and main functions of calculating flood propagation are introduced. Moreover, mathematical model of flood is established, and also the numerical solution and restricted conditions of Saint–venant equations are presented. Using finite difference in solving differential equations is analyzed deeply. As far as Muskingum model be concerned, the important thing is calibrating parameters. Several disadvantages of traditional methodused in calibrating parameters are pointed out, and some optimized algorithms are compared here, including Genetic Algorithm (GA), Annealing Algorithm (SA) and Improved Genetic Algorithm.One of the emphases in this paper is discussing the restricted conditions of Saint–Venant equations and their usage. Meanwhile, the affection on flood propagation, coming from bed roughness, is analyzed also. Researching advanced Muskingum model on long span is another emphases, and a new method is put forward, which divides the long river into several subsections with different distance. After comparing the advantages and disadvantages of each model as well as their applied circumstances, both two models are used in respective situation. The experimental verifications have demonstrated the effectiveness of the proposed methods, which have been applied to the 3D visual simulation system of the Digital Qingjiang Valley.