Study On Basin Flood Model And Its Parameter Inverse Analysis

Study on Basin Flood Model and its Parameter Inverse Analysis(Ph.D. Dissertation)Li GuangchiAbstractChina is a country in which drought and flood has often happened. Water conservancy is play important role in national economy. The development of water conservancy has a direct influence on the stability of nation and the level of people's life. Today, age of science and technology change with each passing day, development of water conservancy must fit for the demand of modernization. Science of water control required a tool of science. Basin flood model is important tools of flood control, promote the beneficial and abolish the harmful, technically water control and scientific water manage. This paper is dealt with that basin flood modeling are compared with the past and sorted out them, a genera! basin flood model is brought forward. The general basin flood model is deduced by analogy between the matter composing of atom and it composing of character objects. The character objects are follow: one dimension flow model, two dimension flow model, zero dimension flow model, linking flow model, engineering control flow model and so on. The different character objects make up of the different basin model. This model contains different flow pattern to fit the demand of different basin model. This model frame is data driver. It runs on existing data. Changed data should automatically change the corresponding model to fit the purpose in general use. Data standardization is the key to realize the model. To make data standard, that different data form drive different methods structure data driver scheme.In order to solve the problem of choice parameter on model, model parameter inverse analysis are discussed in this paper, they include one dimension river flow roughness parameter, tow dimension flow roughnessparameter and side flow of one dimension river flow. One dimension river flow roughness parameter inverse analysis Kalman filter is introduced into the model to solve stochastic error in observed data. Applying Kalman filter automatism revising system, dynamic roughness course is obtained. Using dynamic roughness course the model result precision is improved, it is more consistent with observed data. It is different from traditional concept that the concept dynamic roughness includes influence of model error, truncation error, observation error and so on.In order to analyze side flow of one dimension river flow, the iterative method that comes from control equation system is introduced and applied to the project examples. The results show this method is practicable and is important role in flood control system computation. This method could apply to the inflow computation on large reservoir prevent and control flood.