Research On The Flood Forecasting Model Of Sanshiba Hydropower Station And Its Parameter Uncertainty

Flood forecasting occupies an important position in the non-engineering flood control measures.Predict the flood accurate and timely can effectively assist decision-making on flood control,to reduce or avoid flood damage maximum. Control of reasonable,even may turn the harmful into the beneficial.so, how to reduce the drmage of the flood and Give full play to the economical effect of the reservoirs become the most close attention role.Appropriate use the reservoir for flood control,not only provides guarantee for the water use of downstream ecological and industrial and agricultural production, also offer help for hydropower station to increase the power generation benefit.Therefore carries on the effective and accurate forecasting is an important research, has a strong theoretical and practical significance.There is no system of flood forecasting scheme for Sanshiba reservoir basin,in this paper,the complex Sacramento model is imporved,only keep its runoff yield module and combined with routing model of ARMA,then get a new model base on Sacramento.Make a comparison of the improved project and traditional forecast schemes, there is a greatly enhance to the accuracy and effciency.The main research contents and results of this article are as follows:(1):According to the local hydrological conditions,base on Sacramento model to establish suitable basin hydrological model for local.(2):Through historical flood data of Sanshiba reservoir, the DREAM of statistical algorithm is employed for calibrating model parameters.(3):Make a comparison between the simplified Sacramento model and the experience related to rainfall runoff model,the results show that the simplified Sacramento model forecast effect is superior to experience related to rainfall runoff model.(4)Using GLUE algorithm to analyze the uncertainty of model parameters.