Study On Flood Simulation Of Water-sediment In Daling River Basin

With the increasing development of the society, rivers and reservoirs which are the main water supply sources are increasingly important, which leads to a new high level of water requirement^Sediment forecasting is getting extraordinary necessary with the big damages of rivers and reservoirs caused by sediment.To solve reservoir sedimentation and channel sediment deposition, this paper researches reservoir sediment inflow forecasting, sediment forecasting and sediment process in downstream in Daling river basin. The main content as follows:1. In chapter3, flood events is simulated to predict reservoir sediment inflow. First of all, the relationship between flood influence factors (rainfall, rainfall process, hourly precipitation, flood region composition) and reservoir sediment inflow is analysised; then, sediment forecasting model is built to predict reservoir sediment inflow.2. In chapter4, downstream sediment is predicted by using multiple linear regression model and BP neural network model. In this chapter, sediment in downstream hydrologic station is predicted with the data gathered before Baishi reservoir is built in two different models. The results show that predictive value with BP neural network model is more reliably than predictive value with the multiple linear regression model.3. In chapter5, sediment process is predicted by using similarity theory model. First, similarity of two history floods is analyzed through similarity theory. Then, one flood sediment process of the upstream is used to predict a flood sediment process of the downstream. The result is not good enough for engineering practice because of lack of history flood samples. The model will be more practical in the basin which has enough flood samples.