| With the development of social economy, water is not only the basis for human survival and development, also is one of the important strategic resources in the 21st century, the optimal configuration of water resources become more and more important. As the main measures of regulation of water resources, the flood forecasting of reservoir is necessary for reservoir and downstream of the safety, and reservoir operation plays a vital role in water resources optimization exploitation. At present, as an important river of Liaohe, the flood forecasting and operation way of Chaihe has an important effect. But because the original flood is not very accurate and regardless of the forecast information of scheduling is less effective utilization of water resources. Therefore, we should to research and analyze to determine the better flood forecasting scheme, and based on it make a scheduling approach considering the rainfall forecast information. In this paper, Chaihe as the research background, the main research content is as follows:(1) Using the DHF model and XAJ model based on genetic algorithm (GA) to simulate the reservoir flood. First of all, integrate hydrological data, divide base flow and correct for the 20 floods with consistency, reliability, and representatives. Then use the hydrological forecast model to simulate the floods, optimize parameters based on intelligent algorithm combined with artificial experience. The results show that the floods shape of DHF model based on the infiltration runoff yield and the XAJ model based on filling up runoff yield basically the same as the measured flood, but the simulative floods total slant generally. It is said that the runoff pattern of hydrological forecasting model is not the cause of irrationality. We found that because of the impact of human activities, there has a certain degree of intercept in the basin.(2) Under the consideration of farmland intercept caused by human activity to improve the flood forecast runoff model, in order to enhance the accuracy of flood forecasting. Through field investigation and the applications of remote sensing information analysis, found that there has a certain degree of farmland intercept, and using SPSS software to obtain the factors affecting basin intercept and preliminarily determine the basin intercept function. And using the result to the prediction model to simulate the floods. Through constant adjustment to determine a suitable intercept function. It turned out that the results of XAJ model considering the interception compared with the original, the error is reduced obviously, raise the qualification rate, improve the availability of the forecast results.(3) Based on rainfall forecast information research Chaihe reservoir operation, in order to raise flood control level and ensure that the flood control capacity, at the same time improve the benefits. In level, on the basis of conventional operation, using the rainfall forecast information and flow as indexes to study and develop the flood control forecast operation mode of Chaihe, in order to judging the flood discharge magnitude in advance and changing the way of discharge, and ensure the security of the reservoir and downstream. The results show that the forecast operation based on rainfall forecast information significantly reduce the highest water level compared with the conventional operation, it not only ensure the flood control capacity, but also improves the flood resources utilization. |