基于改进自适应遗传算法的渠系优化配水分析  

Optimization of water distribution in canal system based on improved adaptive genetic algorithm

作　　者：李佳阳 王振华[1,2,3] 张继红 张金珠 刘宁宁[1,2,3] 李淼 胡贵荣 LI Jiayang;WANG Zhenhua;ZHANG Jihong;ZHANG Jinzhu;LIU Ningning;LI Miao;HU Guirong(College of Water Conservancy&Architectural Engineering,Shihezi University,Shihezi,Xinjiang 832000,China;Key Laboratory of Modern Water-saving Irrigation of Xinjiang Production&Construction Group,Shihezi,Xinjiang 832000,China;Key Laboratory of Northwest Oasis Water-saving Agriculture,Ministry of Agriculture and Rural Affairs,Shihezi,Xinjiang 832000,China)

机构地区：[1]石河子大学水利建筑工程学院,新疆石河子832000 [2]现代节水灌溉兵团重点实验室,新疆石河子832000 [3]农业农村部西北绿洲节水农业重点实验室,新疆石河子832000

出　　处：《排灌机械工程学报》2024年第11期1150-1156,1188,共8页Journal of Drainage and Irrigation Machinery Engineering

基　　金：兵团重大科技项目(2021AA003);国家重点研发计划项目(2022YFD1900405);国家重点研发计划项目(2021YFD19008003);国家自然科学基金资助项目(52279040);石河子大学高层次人才项目(RCZK202319)。

摘　　要：为减少渠系在输水过程中的渗漏损失,提高渠系水利用系数,以两级渠系渗漏损失、水流波动最小为目标函数,以下级渠道配水流量、配水开始和结束时间为决策变量,构建配水模型,并采用改进的自适应遗传算法(SIGGA)进行实例求解.提出了一种基于Sigmoid函数的自适应交叉、变异算子,并通过引入信息熵和新种群来减少种群多样性的损失.与人工制定、向量评估遗传算法(VEGA)、多目标粒子群算法(MOPSO)、粒子群混合模拟退火遗传算法(GSPSO)的配水结果进行对比分析,SIGGA求解的配水时长由25.00,15.00,11.00,11.55 d缩短为10.00 d,渗漏损失由97.89×10^(4),71.76×10^(4),70.36×10^(4),32.63×10^(4) m^(3)减少到31.58×10^(4) m^(3),渠系水利用系数由0.651,0.706,0.745,0.828提高到0.833.SIGGA能够克服传统遗传算法(SGA),VEGA易陷入局部最优的缺点,可实现高效、精确地按需配水,为水资源的集约节约化利用提供理论参考.To reduce the leakage loss of the canal system in the process of water transmission and improve the water utilization coefficient of the canal system,a water distribution model was constructed with the minimum leakage loss and minimum flow fluctuation of two-stage canal system as the objective function,the water distribution flow of lower channel and the start and end time of water distribution as the decision variables.The improved adaptive genetic algorithm(SIGGA)was used to solve the problem.A self-adaptive crossover and mutation operator probability based on Sigmoid function was proposed,and the loss of population diversity information entropy was used as an evaluation index,and a new population was introduced to reduce the loss of population diversity.Compared with the water distribution results of manuas formulation,vector evaluation genetic algorithm(VEGA),multi-objective particle swarm optimization algorithm(MOPSO)and particle swarm mixed simulated annealing genetic algorithm(GSPSO),the water distribution duration obtained by SIGGA solution is shortened from 25.00,15.00,11.00 and 11.55 d to 10.00 d,and the leakage loss is reduced from 97.89×10^(4),71.76×10^(4),70.36×10^(4) and 32.63×10^(4) m^(3) to 31.58×10^(4) m^(3),respectively.The water utilization coefficient of the canal system is increased from 0.651,0.706,0.745 and 0.828 to 0.833,respectively.SIGGA can overcome the disadvantage that SGA and VEGA tend to fall into local optimality,and can achieve realize efficient and accurate water distribution on demand,providing theoretical reference for intensive and economical utilization of water resources.

关 键 词：配水模型 自适应遗传算法 SIGMOID函数 信息熵 种群多样性 

分 类 号：S277.9[农业科学—农业水土工程]