Models For The Optimal Allocation Of Water Resources Based On Interval Multistage Stochastic Programming And Applications

With economic development,social development and population growth,the demand for water resources is increasing every year.The over-exploitation of water resources,the water waste and the serious water pollution lead to the shortage of water resources.The optimal allocation model for water resources based on interval multistage stochastic programming is an effective method to solve water shortage and to realize efficient utilization of water resources.However,the former proposed models are difficult to deal with the parameters with interval data of non-normal-distributed stochastic boundaries in the objective functions and constaints.In this thesis,multistage programming is used to describe the dynamic characteristics of water resource management system.Water quantities and economic values are described by interval data with constant boundaries or multiply distributed stochastic boundaries.Two models for the optimal allocation of water resources based on interval multistage stochastic programming are established for two different purposes.Then the water availabilities and each water user's economic values(which contain the economic,social and ecological benefits of water resources)of unit water in Handan city from 2023 to 2025 are given.Finally,the proposed models are applied to water resources management in Handan city.The main contents are as follows:(1)A model for optimal allocation of water resources based on interval multistage stochastic chance constrained programming is given.The model,based on integrating multi-programming,inexact programming and stochastic chance constrained programming,aims for a higher system benefit under the chance constraints of the water quantities(such as flows,allocations and deficits of water).This model is solved by a hybrid algorithm,which incorporates stochastic simulation,multi-layer neural network,and genetic algorithm.Then the feasibility and effectiveness of the proposed model and algorithm are verified by a numerical example of optimal allocation of water resources.(2)A model for optimal allocation of water resources based on interval multistage stochastic dependent-chance programming is given.The model,based on integrating multi-programming,inexact programming and stochastic dependent-chance programming,aims to maximize the probability of the event that the total revenue is not lower than the expected return under the dependent constraints of the water quantities(such as flows,allocations and deficits of water).This model is solved by a hybrid algorithm,which incorporates stochastic simulation,neural network,and genetic algorithm.Then the feasibility and effectiveness of the proposed model and algorithm are verified by a numerical example of optimal allocation of water resources.(3)The water availabilities and the economic values of unit water of four water users(industrial,agricultural,domestic and ecological water users)in Handan city from 2023 to 2025 are given.Firstly,the water availabilities are obtained through the empirical formula of water resources according to the historical information of water quantities in Handan city.Next,with the non-water cost-deducting method and the benefit allocation coefficient method,the losses and the net benefits for industrial,agricultural,domestic and ecological water users when per unit of water is not allocated or allocated are calculated respectively.The water availabilities and each water user's economic values of unit water provide data basis for the application of the two proposed models in Handan city from 2023 to 2025.(4)The application of the examples.The above proposed optimal allocation model for water resources based on interval multistage stochastic programming are applied to the optimal allocation of water resources in Handan city from 2023 to 2025.On the basis of the proposed models,the flows are replaced by the water availabilities.Besides,on the premise of satisfying the domestic water need,the users' priority weights are used to measure the priorities of different users,which are domestic,ecological,agricultural and industrial users in sequence according to the principle of ecological priority.The optimal solutions are obtained by the hybrid intelligent algorithm.The two models provide the optimized configuration schemes of water resources in Handan city for different purposes,reflecting the coordinative development of economic,social and ecological benefits.