基于组合遗传算法的树状管网整体优化设计模型

Tree Pipe Network Optimization Design Method Based on Combined Genetic Algorithm

在灌溉输水系统管网工程的规划设计中,管网布置、水量分配和管径组合是影响管网工程总投资的主要因素。针对树状管网规划优化设计的实际需求,提出了在考虑系统工作制度条件下基于单亲遗传算法和基于整数编码遗传算法的3级优化方法。在用单亲遗传算法对树状管网布置进行优化,得出符合设计要求的若干个管网总长度最短或次最短的布置方案基础上,进一步采用基于整数编码的遗传算法对不同灌溉工作制度条件下管径进行优化,并从经济角度对所选择布置方案的优劣性进行评价,最后选择管网总长度最短且管网投资较小的布置方案作为工程建设采用的最优布置方案。通过工程实例验证分析,提出的组合方法可以进一步使管网系统达到整体的优化,使管网在满足规范设计要求的同时明显降低工程总投资。

In the irrigation water conveyance system pipe network engineering, pipe layout, flow distribution and diameter combina- tion is the most important factor that will influence project investments. The three-step optimal method based on single parent genet- ic algorithm and integer coding genetic algorithm for tree pipe network planning and design is proposed. Single parent genetic algo- rithm to optimize the tree pipe network layout, meet the design requirements of a number o{ the total length of the pipeline network shortest layout scheme. By using integer coding genetic algorithms, the pipe diameters in different irrigation working conditions are optimized, and from an economic point of view, the pros and cons of the selected layout program are evaluated. Finally, the shortest total length of the pipeline network and smaller investment of pipe network layout scheme is selected as the optimal layout scheme of construction projects. The proposed method can further achieve overall optimization of the pipe network, make the pipe network meet the design requirements, while significantly reducing the total project investment.