Resaerch And Application Of Flood Forecasting In Hezhou Hydrological Station Of He River Basin Based On Neural Network

The He River basin has changed because of the production and construction in recent years,in addition,the hydrological department did not accurately predict the flood forecast results in recent three years in the He River basin.Based on the practical work of Hezhou Flood Control and Drought Relief Headquarters Office and in line with the principle of scientific and realistic,the author put forward the topic of Research and application of flood forecasting in Hezhou hydrological station of He River basin based on neural network",and then conducted researching independently.First of all,this paper carried out dimension modeling and data preparation for the business of flood forecast in Hezhou hydrological station,and established a data warehouse.Then,according to the idea of prototype method in software engineering development method and through the analysis of neural network,this improved the forecast program gradually and get a more satisfactory program;Finally,a set of real-time flood forecasting program with a forecast period of 6 hours and a forecasting accuracy of C was obtained.The key point of this research is to design a scientific and reasonable data warehouse for the Hezhou hydrological station in Hejiang River Basin and to design a valuable flood forecasting program so that it can be probed in a more correct direction through the analysis of the neural network and the test of the actual data.Establishing data warehouse on flood forecasting,the application of data mining and neural network to flood forecasting in Hezhou were the new breakthroughs in flood forecasting in Guangxi,and it will be optimized continuously and applied to other hydrological station for greater achievement.