| Nowadays, flood is one of the main natural disasters in the world. So how to control the flood and reduce the effect of flood is the general problem which has been paid much attention to throughout the world. The accurate flood forecast is the precondition and important guarantee of conquering the flood and reducing the disaster loses. While the accuracy of flood forecast mainly lies on the configuration and parameters of the model. After choosing a proper model configuration according to the idiographic condition of the drainage area, the problem of choosing model parameters is more important, so seeking a parameters comprehensive optimized method has important academic and practical meaning to improve the accuracy of flood forecast and realize the optimize system conttol.This article is on the basis of the project on automated measuring and recording water condition and flood prediction system of Da Ying River, and reaches its core----flood forecast. Contraposing the problem that estimating parmeters with the trail-and-error calibration method is difficult and the accuracy is low,I introduce PSO ,Making the most use of the hydro meteorological data of the Da Ying River, I study the academic method of optimizing Xinanjiang model parameters and Tank model parameters with PSO,analyze and compare the results.If we compare the parameters optimizing method through the qualified rate and confirmation modulus of historical floods, we can find that the accuracy of optimizing Xinanjiang model parameters and Tank model parameters with PSO is higher than manual calibration,so it shows the advantages of PSO。At the same time, I compared Xinanjiang model and Tank model after optimizing parameters with PSO, the accuracy of the former one is higher than the latter one. So I think we would better take Xinanjiang model as the main forecast model and Tank model as the assistant model in Da Ying River area automated measuring and recording water...
|
PDF Full Text Request