Study On Water Demand Prediction And Water Allocation In Kezi River Basin

Water is the base of human existence and social economy development. The surface water exists in the form of runoff. Many factors could influence the formation of surface runoff,such as human activities and natural changes, so the hydrological characteristics are very complex. Therefore, the establishment of runoff prediction model based on the characteristics and regularity analysis of runoff time series could improve forecast accuracy and provide guidance for reservoir operation. In China, water resource distribution is unbalanced and the per capita occupancy volume is low. The basic national conditions determine the importance of water resources in agricultural production and the economic and social development. The 2015 No.1 document states that there exist the problems of agricultural resources shortage, serious pollution, overexploitation in China, improving our country’s agricultural resources sustainable development ability f irst to solve the problem is guarantee agricultural economic development, improve the utilization rate of agricultural resources, strengthen groundwater over mining area and the comprehensive management of agricultural water metering and establish agricultural quota management and irrigation water volume control system under the constraint of resources and environment protection. Water reasonable allocation is the key to total amount control of water supply. it promote the development and utilization efficiency of water resources and has very important practical significance.In this paper, with the Kezi River Carla Bailey reservoir and Kashi as the research object, mainly study the following three aspects:(1) Study on wavelet variance with matlab to analysis the data of trends and different period. It provides basis for runoff forecasting. Then we established an annual runoff model of BP neural network under different period. The study proved that BP neural network prediction model based on the analysis of wavelet variance with higher accuracy.(2) With Carla Bailey mountain reservoir as an example, using genetic algorithm to establish the optimization model of reservoir operation which combined with the industry water demand, agriculture water demand, domestic water demand, to meet the reservoir irrigation guarantee rate and power guarantee rate, then use the MATLAB programming language to solve the model.(3) Through analysis the situation of supply and demand of agr icultural water resources in Xinjiang Kashi Economic Zone, combined with irrigation systems, crop types and irrigation schedule, offering proposals of basin surface water-groundwater resource under the water supply limitation, distributing the amount of water supply to the township of Kashi City.