Classifying Of Complicated Datum And Study On Application Of EV Model

The lower reaches of the YelowRiver belong to scouring and silting channel, because of water-sedimentraiting curve , channel's boundary and their interaction, the changing process of stage is complexity, stage's fluctuation is often quite great in the almost same runoff. Contraposing the datum trait of the lower reaches of the YelowRiver, advancing the ideas of classifying the datum and building model. According to the relations of stage, runoff and sediment concentration, advancing combination-clustering analysis method and presenting a new rule of classifying. It show that the method can classify effectively by comparing the relative coefficient and results of stage forecasting. Building the stage forecasting model and presenting the better results of forecasting. As there are errors in measured datum, EV model which more accord with the fact duing to taking into account the errors of the independent are introduced into estimating parameters. Presenting the iterative formula of estimates of parameters of non-linear EV model based on the maximum-likelihood estimation when the errors of the independent and the dependent are independent. Introducing into a easier arithmetic given by Hubert Schwetlick and Volker Tillerfor EV model and mending it, presenting its MATLAB procedure applied on the stage forecasting model. Exploring the application of EV model in the stage forecasting model.