River flood forecasting method based on flood type generalization

Abstract

The invention provides a river flood forecasting method based on flood type generalization. History data of a forecasting site is analyzed; a typical flood hydrograph of the site is determined; forecasting factors of flood peak / duration of the site are selected; a forecasting model of the flood peak / duration is established through adoption of a support vector machine; a typical flood mode is zoomed according to period-by-period flood peak / duration forecasting results of the model; and further period-by-period flood forecasting results of the site are obtained. Compared with the existing methodof forecasting a whole process of the flood, the method has the advantages that flood body shapes, peaks and duration are emphatically forecasted, a typical flood process of the to-be-forecasted siteis determined through utilization of a flood type generalization method, the forecasting model of the flood peak and duration is established through adoption of a support vector machine algorithm, finally a typical process is zoomed through utilization of the flood peak and duration forecasting results, and a whole flood forecasting result is obtained. Compared with the existing hydrology methodand a water conservancy model, the method is characterized by simplicity and rapidness.

Description

technical field [0001] The invention relates to a river course flood forecasting method, in particular to a river course flood forecasting method based on flood type generalization. Background technique [0002] River flood forecasting has always been an important and difficult point in flood control and drought relief work, and reasonable forecast results are of great significance for Xingli flood control. At present, there are mainly two kinds of river flood forecasting methods, namely hydrological method and hydraulic method. Among them, the hydrology method is the comprehensive method of the Muskingum method, the Muskingum water level simulation method and the diffuse wave nonlinear water level method; while the hydraulic method uses iterative solution to the four-point eccentric implicit scheme discrete Saint-Venant equations method. [0003] However, due to the strong empirical parameters of the hydrological method, it is difficult to accurately calibrate the paramet...

AI Technical Summary

Problems solved by technology

[0003] However, due to the strong empirical parameters of the hydrological method, it will be difficult to accurately calibrate the parameters in the watershed with complex flow conditions; and the hydraulic method has the disadvantages of requiring more data of different types and long calculation time.

Images