Research And Application In Automatic Calibration Of Hydrologic Model Parameters Towards Separated Floods

Parameters of hydrologic models often not exactly known and therefore have to be determined by calibration.A manual calibration depends on the subjective assessment of the modeler and can be very time-consuming though.Method of automatic calibration can improve these shortcomings.However,this is a kind of multidimensional and nonlinear problem,and always include some Multi-objective decision-making difficulty,so we mainly focus on the Xin'anjiang model parameters optimazition and flood forecast calibrating system design issues.We have get relative Satisfactory results.The main research coments are as following:(1)According to the historical data of ShangPai reservoir in year between 2000 and 2007,we divide the flow course into 45 floods,considering the recede course and the flow course of ChenTianHe reservoir.We use 34 floods to caliberate the model parameters,and 11 floods to check the results,we choose flood peak value,volumn and peak appear time as the object functions and the fuzzy muti-objective optimization is developed to solve the muti-objective optimazition problem.(2)According to basic information of ShuangPai reservoir in HuNan province,this paper use SCE-UA and the PSO algorithm calibrating the Xin'anjiang model parameters,and compared the two methods of their respective advantages and disadvantages.The results showed that the peak and volume of flood reached a Class-A standards,and the peak time reached more than a grade B standard,the Satisfactory results shows that the two algorithms are entirely in the actual use of flood forecasting.(3) Through the research and analysis towards the characteristics of ShuangPai basin,the web-based ShuangPai Reservoir One-hour Flood Forecast Calibrating System was completed.The system is technologically advanced and reliable,safe and easy to use.it provide a distributed computing platform and can calibrate parameters based on Internet environment.We achieved good results by joint application of this system and the original flood forecasting and dispatching system.