基于粒子群算法的给水管网优化设计

摘要

Pipe diameters of water supply network are designed by particle swarm algorithm optimization to achieve the goal of lowest annual conversion cost of network. Penalty function is set for each node to ensure that pressure of each node satisfy the constraint of minimum pressure. Pipe diameters searched by the algorithm in each step are rounded to be standard to ensure that final combination scheme should meet the requirement of engineering. Comparing with genetic algorithm which is widely used, the PSO reduces computational complexity, not only has rapid convergence ability but also owns good global searching ability.%以管网造价年费用折算值最低为目标,采用粒子群算法对给水管网进行优化设计;首先对节点压力设置罚函数,对不满足压力要求的节点压力进行惩罚,以保证管网中各个节点都满足最小压力约束;由于工程中所用管径值全部都是离散值,算法对搜索得到的结果进行标准化处理,保证求得的管径组合方案符合实际要求;与给水管网优化设计中常用方法遗传算法相比,粒子群算法降低了计算复杂度,具有较好的求解性能,而且具有较快的收敛速度和全局搜索能力.