基于多目标粒子群算法的渠系优化配水研究  

Optimizing Water Distribution in Canal Networks Using Multi-objective Particle Swarm Optimization Method

作　　者：李彤姝 黄睿 孙志鹏 郭珊珊[1] 易康 韩宇[1] 陈建[3] LI Tongshu;HUANG Rui;SUN Zhipeng;GUO Shanshan;YI Kang;HAN Yu;CHEN Jian(College of Water Resources&Civil Engineering,China Agricultural University,Beijing 100083,China;College of Information and Electrical Engineering,China Agricultural University,Beijing 100083,China;College of Engineering,China Agricultural University,Beijing 100083,China)

机构地区：[1]中国农业大学水利与土木工程学院,北京100083 [2]中国农业大学信息与电气工程学院,北京100083 [3]中国农业大学工学院,北京100083

出　　处：《灌溉排水学报》2020年第9期95-100,125,共7页Journal of Irrigation and Drainage

基　　金：国家自然科学基金项目(51979275);国家重点研发计划项目(2017YFD0701003);吉林省重点科技研发项目(20180201036SF);中国农业大学基本科研业务项目(10710301)。

摘　　要：【目的】发展节水高效农业,实现对农田渠系配水的优化。【方法】以水流过渡平稳和输水损失最小作为优化目标,采用多目标粒子群算法建立渠系优化模型,以西北地区西浚灌区干、支两级渠系作为研究对象进行了模型求解,并与回溯搜索法和向量评估遗传算法求解结果进行了比较。【结果】采用多目标粒子群算法得到的渠系配水时间缩短至约11 d,配水总量为205.07万m^3,较回溯搜索法和向量评估遗传算法求解得到的渠系配水时间短2~4 d。【结论】多目标粒子群算法在满足双优化目标的基础上,能够贴近实际配水情况并合理地调整流量的时空配置。【Background】Most irrigation districts use canal network to distribute water and optimizing water distribution process in the network can improve irrigation water use efficiency and reduce water waste.【Objective】The purpose of this paper is to present methods to optimize water distribution in canal network to help achieve multi-scale optimal allocation of water resource in developing water-saving agriculture.【Method】We took steady flow and minimum water loss as the optimization objectives,and used the multi-objective particle swarm optimization(MOPSO)algorithm to establish a multi-objective canal system optimization model.The model was solved by the two-level canal system with Xijun Irrigation District taken as an illustrative example.The results were compared with that obtained from the backtracking search algorithm and vector evaluation genetic algorithm(treated as CK).【Result】The results revealed that the irrigation period obtained from the MOPSO for the canal network was about 11 days less,with the total water being 205.07 million m^3.Compared with the CK,the irrigation period calculated by the proposed method in the canal system was 2~4 days less.【Conclusion】The MOPSO is an effective method to calculate water distribution in the canal network.The calculated results matched the measured water distribution well,and it can be used to manage water distribution both spatially and temporally to satisfy the two optimization objectives.

关 键 词：渠系 优化配水 多目标约束 粒子群算法 

分 类 号：S274.3[农业科学—农业水土工程]