摘要
【目的】探究粒子群算法和天牛群算法在求解灌区渠系配水模型时的性能差异及优化后配水方案的共有特性。【方法】以大功灌区总干、分干两级渠系为研究对象,依据灌区不同的用水情形将其划分为18个配水对象,以灌溉用水总量、渠系渗漏水量及总干渠流量波动大小为优化目标,以下级渠道输水流量和输水启、闭时间点为决策变量,构建多目标两级渠系配水模型,分别使用粒子群算法和天牛群算法对模型进行求解,基于求解结果结合熵权-TOPSIS法对2种算法的性能进行综合评价。【结果】熵权-TOPSIS法的评价结果表明,粒子群算法的整体性能优于天牛群算法,但后者的计算速度高于前者。此外,同一用水情形下2种算法求解的配水方案相近,且优化后的下级渠道配水流量与配水时长存在共性规律。【结论】“左右须”寻优机制的引入使天牛群算法的计算速度相比粒子群算法最高可提升56%,但由于“左右须间距”初始参数的设置问题,随着计算维度的增加,粒子群算法的整体性能优于天牛群算法。研究结果可为灌区渠系配水管理提供科学依据。
【Objective】To investigate the performance difference between the Particle Swarm algorithm and the Beetle Swarm Optimization algorithm in solving the water distribution model of a two-level canal system in irrigation districts,as well as the common characteristics of the optimized water distribution schemes.【Method】This study takes the two-level canal system of Dagong Irrigation District as the research object,and divides it into eighteen water distribution scenarios according to different water use situations in the irrigation district.The total amount of irrigation water,the amount of water leakage and the fluctuation of the flow rate of the main canal are used as the optimization objectives.The decision-making variables are the flow rate of the sub-main canals and the opening and closing time points of the water transmission.Construct a multi-objective two-level canal distribution model,solve it using the Particle Swarm Optimization algorithm and the Beetle Swarm Optimization algorithm respectively,and comprehensively evaluate the performance of the two algorithms based on the solution results combined with entropy weight-TOPSIS method.【Result】The evaluation results of entropy weight-TOPSIS method show that the performance of the Particle Swarm Optimization algorithm is better than the Beetle Swarm Optimization algorithm,but the computational speed of the latter is significantly faster than that of the former.In addition,the water distribution schemes solved by the two algorithms under the same water use situation are close to each other,and there is a common law between the optimized water distribution flow and duration of the sub-main canal.【Conclusion】The results of the study can provide suggestions for the management of water distribution in two canal systems in irrigation districts and provide a basis for selection algorithms.
作者
卢德友
田桂林
杨泊
秦京涛
LU Deyou;TIAN Guilin;YANG Bo;QIN Jingtao(Henan Vocational College of Water Conservancy and Environment,Zhengzhou 450011,China;Farmland Irrigation Research Institute,Chinese Academy of Agricultural Sciences/Key Laboratory of Water saving Irrigation Engineering,Ministry of Agriculture and Rural Affairs,Xinxiang 453002,China;Graduate School of Chinese Academy of Agricultural Sciences,Beijing 100081,China)
出处
《灌溉排水学报》
CAS
CSCD
北大核心
2023年第9期129-137,共9页
Journal of Irrigation and Drainage
基金
2023年度河南省高等学校重点科研项目(23B570003)
中国农业科学院科技创新工程项目(ASTIP)。
关键词
灌区
渠系配水
天牛群算法
粒子群算法
熵权-TOPSIS评价模型
irrigation district
canal system water distribution
the particle swarm optimization algorithm
the beetle swarm optimization algorithm
entropy weight-TOPSIS evaluation model