[1]李轩,侯精明,同玉,等.基于代理模型辅助PSO算法的排水管网优化设计[J].中国给水排水,2022,38(17):117-124.
　LIXuan,HOUJing-ming,TONGYu,et al.Optimization Design of Drainage Network Based on Surrogate Model Assisted Particle Swarm Optimization Algorithm[J].China Water & Wastewater,2022,38(17):117-124.
点击复制

基于代理模型辅助PSO算法的排水管网优化设计

中国给水排水[ISSN:1000-4062/CN:12-1073/TU] 卷: 第38卷 期数: 2022年第17期 页码: 117-124 栏目: 出版日期: 2022-09-01
Title:
Optimization Design of Drainage Network Based on Surrogate Model Assisted Particle Swarm Optimization Algorithm
作者:
李轩1， 侯精明1， 同玉1， 杜颖恩1， 薛树红2
（1.西安理工大学 省部共建西北旱区生态水利国家重点实验室， 陕西 西安 710048；2.中国电建集团西北勘测设计研究院有限公司，陕西 西安 710065）
Author(s):
LI　Xuan1， HOU　Jing-ming1， TONG　Yu1， DU　Ying-en1， XUE　Shu-hong2
（1. State Key Laboratory of Eco?hydraulics in Northwest Arid Region of China， Xi’an University of Technology， Xi’an 710048， China； 2. PowerChina Northwest Engineering Corporation Limited， Xi’an 710065， China）
关键词:
城市雨洪;  排水管网;  代理模型;  粒子群优化;  SWMM
Keywords:
urban flood;  drainage network;  surrogate model;  particle swarm optimization;  SWMM
摘要:
近年来城市暴雨洪涝灾害日益严重，设计良好的排水管网则是缓解城市洪涝最为经济有效的方式之一。但目前排水管网一般采用经验方法进行设计，不仅设计成本较高，而且管网排水能力有限。基于此，利用程序将SWMM模型嵌入粒子群算法寻优过程，以管道溢流量最小为目标，并以经济成本和水力特性为约束条件建立优化模型；同时，为了克服粒子群算法最优解波动较大和计算效率较低问题，通过耦合Kriging模型，提出了一种基于在线代理模型辅助粒子群算法的排水管网优化设计方法。以西安市某排水分区为例进行验证，并与常规的设计方法进行对比，结果表明，所提出的方法可在满足约束条件下实现洪涝缓解效果的最大化，相比于规划设计方法，管道溢流量降低了33.35%。此外，与标准粒子群算法优化结果相比，平均计算时间减少了27.56%，且优化效果更为显著。
Abstract:
In recent years, urban flood disasters have become increasingly serious. A well-designed drainage network is one of the most economical and effective ways to alleviate urban floods. However, empirical methods are generally adopted to design the current drainage network, which not only has a high design cost, but also has limited drainage capacity. Therefore, SWMM was embedded into the optimization process of particle swarm algorithm by program modification, and an optimization model was established with economic cost and hydraulic characteristics as constraints to minimize the pipeline overflow. In addition, a drainage network optimization design method based on online surrogate model assisted particle swarm optimization algorithm was proposed by using coupling Kriging model to overcome the large fluctuation and low computational efficiency of the optimal solution of particle swarm optimization algorithm. The design method was verified in a drainage district in Xi’an and compared with the conventional design method. The proposed method maximized the flood mitigation effect under the constraints, and the pipeline overflow was reduced by 33.35% compared with the planning scheme.In addition, compared with the optimization results of the standard particle swarm optimization algorithm, the average computation time was reduced by 27.56%, and the optimization was more significant.

相似文献/References:

[1]杨东,侯精明,李东来,等.基于扩散波方法的管网排水过程数值模拟[J].中国给水排水,2020,36(15):113.
　YANG Dong,HOU Jing-ming,LI Dong-lai,et al.Numerical Simulation of Pipe Network Drainage Process Based on Diffusion Wave Method[J].China Water & Wastewater,2020,36(17):113.
[2]陆露,高峰,郭娟,等.排水管网运维管理问题分析与对策研究[J].中国给水排水,2022,38(2):8.
　LU Lu,GAO Feng,GUO Juan,et al.Problem Analysis and Countermeasure Research for Sewer Operation and Maintenance Management[J].China Water & Wastewater,2022,38(17):8.
[3]邵宇航,楼少华,唐颖栋,等.深圳市茅洲河流域某小微水体治理方法与实践[J].中国给水排水,2023,39(14):134.
　SHAOYu-hang,LOUShao-hua,TANGYing-dong,et al.Method and Practice of a Miniature Water Body Management in Maozhou River Basin of Shenzhen[J].China Water & Wastewater,2023,39(17):134.

更新日期/Last Update: 2022-09-01