Research On The Optimization Method Of Water And Land Resources In The "Reservoir And Pumping Station" Irrigation System Under Water Supply Restriction

To improve the utilization efficiency of agricultural resources,it is necessary to optimize the allocation of limited irrigation water and crop distribution simultaneously.The study on the optimization method of water and land resources for the joint operation of multi-water source projects under the condition of limited water supply is of great significance for improving the efficiency of agricultural water use and ensuring the sustainable development of regional economy.Aiming at the complex irrigation system of irrigation reservoirs(groups)and pumping stations(groups)in water-scarce areas,the optimal allocation of water resources and the optimal layout of crop planting are studied in this paper,key research findings include:(1)A method for optimizing the allocation of water and land resources in a "single reservoir and single pumping station" irrigation system.Taking a single annual regulating irrigation reservoir with a single pumping station as the research object,first considering the situation of a single dry crop,with the maximum output value of a single crop as the objective function,the water supply and disposal volume of the reservoir at each time period,and the lifting volume of the replenishment pump station at each time period as decision variables,and the total water supply volume of the reservoir and scheduling criteria constraints as constraints,a "single reservoir and single pumping station" irrigation system water resource optimization allocation model under water supply constraints is constructed.Considering that the model only contains one-dimensional state variables,the corresponding one-dimensional dynamic programming(DP)recursive method is used to solve the problem.Taking Yibei Irrigation District in Xinyi City,Jiangsu Province as an example,the application results show that the increase in relative crop yield is not only affected by the optimization and adjustment of irrigation water supply distribution,but also by a decrease in system evaporation loss and an increase in irrigation water volume.In drought years and special drought years,DP increased output value by 8.9%and 3.6%compared to standard operation policy(SOP),respectively;Compared to genetic algorithm(GA),it has increased output value by 1.0%and 1.4%,respectively.A "single reservoir and single pumping station" irrigation system water resource optimization allocation model based on rotation of two main crops has been established.The model takes the maximum total output value of the irrigation area as the objective function,and takes the water supply and disposal volume of the irrigation reservoir at each stage,as well as the water replenishment volume of the replenishment pumping station at each stage as the decision variables.The total annual water supply volume of the reservoir,the total annual water withdrawal volume of the pumping station,and scheduling criteria are constraints.In view of the complexity of the model,which only contains one-dimensional coupling constraints,the large system decomposition aggregation dynamic programming method(DADP)is used to solve the model,the comparison with various heuristic algorithms shows that the DADP algorithm has good algorithm performance.Finally,based on the "single reservoir and single pumping station" irrigation system water resource optimization allocation model based on crop rotation,the crop planting area is also used as a decision variable to establish the "single reservoir and single pumping station" irrigation system water and land resource optimization allocation model.The model can be transformed into a high-dimensional dynamic programming model when the area decision variables are determined;When determining the water supply and water spill of the reservoir,and water water replishment the pumping station,it can be transformed into a one-dimensional linear model.Therefore,a combination algorithm((DADP-LP)SA)of decomposition aggregation dynamic programming method(DADP)and linear programming(LP)successive approximation solution is proposed.The comparison between this algorithm and the heuristic algorithm shows that the combination algorithm also has good algorithm performance.(2)A method for optimizing the allocation of water and land resources in a "multi reservoir and multi pumping station" irrigation system.Take multiple irrigation reservoirs and multiple pumping stations as the research object.The objective function is to maximize the total output value of irrigation areas supplied by multiple reservoirs,with the water supply,water spill of the reservoir,and water replenishment of each pumping station at each stage as decision variables.The total available water supply of the "multi reservoir and multi pumping station" large-scale system,the annual available water supply of each reservoir,the annual available water supply of each pump station,and the scheduling criteria of each reservoir are constraints.Based on crop rotation,a water resource optimization configuration model for the " multi reservoir and multi pumping station" irrigation system with limited water supply is constructed.Considering the hydraulic connection between reservoirs,a multi level decomposition aggregation dynamic programming method(MDADP)for large-scale systems is proposed for the first time.Apply the model and method to Yibei Irrigation District and Gao’a Irrigation District in Xinyi City,Jiangsu Province,and compare and analyze the algorithm’s optimality and applicability with modern heuristic algorithms.The results show that compared to heuristic algorithms,MDADP can obtain better objective function values,converge faster,have higher algorithm stability,and be easy to operate,with better algorithm optimality;At the same time,analyzing the scheduling process and results of irrigation systems,MDADP is also more applicable.Similarly,taking crop planting area as a decision variable,a water and land resource optimization allocation model is constructed based on the parallel "multi reservoir and multi pumping station" irrigation system water resource allocation model.According to the characteristics of the model,a combination algorithm of multi-level decomposition aggregation dynamic programming algorithm(MDADP)and linear programming(LP)successive approximation(MDADP-LP)SA is proposed.Compared with heuristic algorithms,the results show that although the combination algorithm has the disadvantage of long solving time,it exhibits good algorithm performance in terms of objective function optimality and algorithm stability,providing a new solution approach for planning solutions similar to complex high-dimensional models.