基于 PSO 原理的异孔径配水系统优化算法

摘要

为了实现穿孔配水系统均匀布水的快速优化设计，提出采用改进的粒子群优化算法(PSO)并结合水力模型工具作为系统均布优化决策制定的方法。该算法首先对常规粒子群算法进行了一定的改进，用于提高其算法的收敛性及收敛的速度；其次将改进后 PSO 与水力模型通过外部 TOOLKIT 接口进行整合，以模型计算后数值作为集群个体适应度的评估指标。算例研究表明：该方法可有效地实现限定条件下的均布配水系统快速优化设计。同时，结合配水系统的节点压力分区，可减小粒子群算法的维度，从而使算法的收敛速度得到较大的提高，迭代次数减少了83%，并能产生更符合工程实际的优化方案。%To achieve optimal design of a perforated water distribution system, this paper proposes a particle swarm optimization ( PSO) algorithm combined with hydraulic modeling tools as a systematic approach to decision making. First, the algorithm improved the standard PSO to enhance the convergence rate and the capability of global convergence. Then, with the external TOOKKIT interface, a combination of the improved PSO with hydraulic models had been developed for fitness evaluation. The case study shows that this method is capable of realizing optimization of the design program with certain limited conditions. Meanwhile, a scheme corresponding to engineering practice can be developed with the strategy of node pressure grouped, and the approach reduces the particle dimensions and speeds up the convergence rate about 83% .