Water Demand Forecast And Water Resources Optimization Of Lingyuan City

Water resources are precious resources and indispensable material conditions for human survival,social progress and economic development.It is also an essential element for improving the ecological environment.With the development of industrial and agricultural production and people's pursuit of high quality of life,the demand for water resources in quantity and quality has also increased.The contradiction between the supply and demand of water resources has become increasingly acute and now has become a“bottleneck”,that restricts the development of the national economy.Therefore,the establishment of a rational and applicable water resource optimization configuration model has important guidance and practical significance for the local comprehensive development.This paper was based on the analysis of the theories and methods of water demand forecasting and optimal allocation of water resources at home and abroad.Taking Lingyuan City as the research area,a water demand forecasting model based on principal component analysis(RBF)neural network and a water resources optimization model based on genetic algorithm was established.The main research results are as follows:(1)According to the status of surface water and groundwater supply of Lingyuan City in2015,and on the basis of assessing the daily water supply of reuse and new water supply projects before 2020.It is predicted that the water supply for surface water,groundwater,reuse of reclaimed water,and new water supply projects of Lingyuan City's in 2020 will be19.04 million m~3,85.01 million m~3,23.1 million m~3,84.4 million m~3.(2)A water demand prediction model of RBF(PCA-RBF)neural network based on principal component analysis was established,and compared with the RBF neural network demand forecasting model.The result shows that:The relative error of PCA-RBF neural network water demand prediction model is 2.9%and 0.4%,The relative error of RBF neural network water demand prediction model is 2.3%and 0.3%.The structure of PCA-RBF neural network is 5×3×1,and the structure of the RBF neural network is 13×8×1.This shows:The prediction accuracy of the two models is basically the same.The PCA-RBF prediction model is more simplified,the training is relatively easy.Moreover,it has less application demand for future prediction data,and has a good application prospect and wide range of prediction.(3)Based on the theory of sustainable development of water resources,a multi-objective optimal allocation model of water resources has been established.Analyze and determine the parameters of the model,then the large-scale system optimization method based on genetic algorithm is used to solve the model.The results of optimal allocation of water resources in Lingyuan in 2020 were obtained,and the allocation of water resources under the condition of agricultural water saving was also considered.The water scarcity rates of the two options are9.30%and 4.67%.The coefficient of regional coordinated development is greater than 0.9,indicating that Lingyuan City is in a coordinated development.The research results can provide theoretical and practical reference for the sustainable development of water resources in Lingyuan City.