基于多目标混合蛙跳差分进化算法的皂市水库优化调度  

Optimal Operation of Zaoshi Reservoir Based on Multi-objective Shuffled Frog Leaping Differential Evolution Algorithm

作　　者：杨丽娟 薛联青[1,2] 伍佑伦 盛东 刘远洪 李娜[3] 周天文 何怀光 YANG Li-juan;XUE Lian-qing;WU You-lun;SHENG Dong;LIU Yuan-hong;LI Na;ZHOU Tian-wen;HE Huai-guang(College of Hydrology and Water Resources,Hohai University,Nanjing 210098,Jiangsu Province,China;School of Hydraulic Engineering,Wanjiang University of Technology,Maanshan 243031,Anhui Province,China;Hunan Institute of Water Resources and Hydropower Research,Changsha 410114,Hunan Province,China)

机构地区：[1]河海大学水文水资源学院,江苏南京210098 [2]皖江工学院水利工程学院,安徽马鞍山243031 [3]湖南省水利水电科学研究院,湖南长沙410114

出　　处：《中国农村水利水电》2023年第11期27-34,44,共9页China Rural Water and Hydropower

基　　金：国家自然科学基金项目(51779074);湖南省水利科技项目重大项目(XSKJ2021000-10);江苏省水利科技项目(2019031)。

摘　　要：随着社会经济的发展和对生态保护需求的增强,以防洪和兴利为目标的传统水库综合调度已经不能满足多样化的需求,进行多目标的水库综合优化调度显得尤为重要。为实现皂市水库水资源综合效益最优化,平衡社会效益、经济效益和生态效益,面向水库的发电、供水和生态要求,在不同来水条件下建立基于多目标混合蛙跳差分进化算法(Multi-objective Shuffled Frog Leaping Differential Evolution Algorithm,MSFL-DEA)求解的皂市水库优化调度模型,并采用层次分析法(Analytic Hierarchy Process,AHP)与逼近理想点法(Technique for Oder Preference by Similarity to Ideal Solu⁃tion,TOPSIS)的组合赋权法进行不同决策偏好的方案优选。结果表明:随上游来水减少,水库的发电效益下降最为显著,其次为生态效益和供水效益;各来水频率下,生态与发电、供水与生态的竞争关系相对更为明显,发电与供水之间相对较弱;根据不同决策偏好,得出供水优先、发电优先、生态优先以及均衡优化4组方案,对比认为均衡方案偏向折中,且年内各月的供水、生态及发电过程也相对协调(以P=75%为例);该方案在来水频率分别为50%、75%和95%时,年发电量分别为3.697、2.981、2.155亿kWh,供水满足度为0.871、0.775、0.702,生态满足度为0.861、0.699、0.615。研究成果有利于提高水资源利用效率,为皂市水库的科学调度决策提供参考。With the development of society and the increasing demand for ecological conservation,the traditional integrated reservoir opera⁃tion aiming at flood control and profit fail to meet the diversified requirements.Hence,the implementation of multi-objective reservoir com⁃prehensive optimization scheduling has become increasingly urgent.In order to achieve the optimization of the comprehensive benefits of Za⁃oshi Reservoir and balance the social,economic and ecological benefits,the optimization dispatching model of Zaoshi Reservoir for power generation,water supply and ecological requirements is established,which adopts the improved Multi-objective Shuffled Frog Leaping Dif⁃ferential Evolution Algorithm(MSFL-DEA)to calculate.And the scheme based on AHP and TOPSIS combined weight method is optimized according to different decision preferences.The results show that with the decrease in upstream inflow,the power generation benefit of the reservoir decreases most significantly,followed by the ecological benefit and water supply benefit.Under each inflow frequency,the competi⁃tion relationship between ecological goal and power generation goal,water supply goal and ecological goal is more obvious,and the competi⁃tion between power generation goal and water supply goal is relatively weak.Depending on decision preferences,four schemes of water sup⁃ply priority,power generation priority,ecological priority and balance optimization are obtained.By comparison,it is found that the bal⁃anced scheme tends to compromise,and the water supply,ecology and power generation processes are relatively coordinated in each month of the year(P=75%as an example).When inflow frequency is 50%,75%and 95%,the annual power generation is 3.697,2.981 and 2.155 billion kWh,the water supply satisfaction is 0.871,0.775 and 0.702,and the ecological satisfaction is 0.861,0.699 and 0.615,respective⁃ly.The research results are conducive to improving the efficiency of water resources utilization,and can provide reference for the scientific

关 键 词：皂市水库 多目标优化调度 供水 发电 生态 

分 类 号：TV697[水利工程—水利水电工程]