Research On Two-Stage Optimal Model And Algorithm Of Tree Irrigation Pipe Network

The irrigation of water-delivery with pipelines is an important way to develop the water-saving agriculture, and its tendency is the pipeline irrigation system. In the whole pipe irrigation network system, the pipe network investment generally occupies 50-80% of the total project, moreover, it involves huge energy consumption and operational management costs. Therefore, in the situation of the limited project fund, it has practical significance to perform the optimal design for saving the investment, reducing the energy consumption, enhancing the economic and social benefits. The guiding principle of optimal design is to satisfy the water volume and irrigation pressure request, seek the lowest cost or the smallest yearly operation expense and have the highest system reliability.The research object is tree pipe network. Based on the summarization of the domestic and foreign pipe network optimization and the optimized technology achievement, in view of the existing problems and shortcomings, this paper studies the tree irrigation pipe network layout and diameter design optimization, combining theoretical study and optimization algorithm with computer analysis and case analysis, adopting Genetic Algorithm(GA) optimization theory and Graph Theory(GT). This paper develops an optimal design software with the aid of Visual Basic 6.0 and MATLAB, using mix programming technology, that is more general to tree irrigation pipe network and relies less on the experience of the designers.This paper mainly achieved the following results:(1) This paper studies tree pipe network focused on the farmland irregular irrigation pipe network. The model eliminated the redundant links so that each node is supplied only by one single link. In the first step, the pipe network is simplified as a directed network graph, which rejects the complexity coding method in GT. A simple coding method is introduced, that improves the calculation efficiency of GA, and enhances the operation of optimization method.(2) According to the design experience in the first-step layout optimization design, this paper determines all the possible water supply pipes of each node. Meanwhile, the design experience is organically applied in the beginning of the optimization method, which reduces the number of infeasible solutions and enhances the calculation efficiency and feasibility.(3)Using the Genetic Algorithm Toolbox which Sheffield University programmed, this paper applies the integer coding method in Genetic Algorithm programming and avoids redundant problems of binary coding with the help of MATLAB. The interface of this application software is programmed by VB and has the visual operation method. The network feature data can be inputted in the form of corresponding matrix. The software can be packaged and run in any computer with MATLAB.