基于混合多目标粒子群算法的梯级橡胶坝群蓄洪调度研究  

Research on Flood Storage and Scheduling of Cascade Rubber Dam Groups Based on HMOPSO Algorithm

作　　者：徐伟[1] 臧旭东 夏冰 张磊 杨蕾 XU Wei;ZANG Xudong;XIA Bing;ZHANG Lei;YANG Lei(College of Water Conservancy,Shenyang Agricultural University,Shenyang 110161,China;Fuxin Water Conservancy Survey and Design Institute Co.,Liaoning Ltd.,Fuxin Liaoning 123099,China)

机构地区：[1]沈阳农业大学水利学院,沈阳110161 [2]阜新市水利勘测设计研究院有限公司,辽宁阜新123099

出　　处：《沈阳农业大学学报》2025年第1期73-81,共9页Journal of Shenyang Agricultural University

基　　金：辽宁省教育厅课题项目(JYTZD2023124)。

摘　　要：[目的]城市段河道梯级橡胶坝群蓄洪调度问题具有非线性、多维性和高约束性,研究其高效稳定的求解方法,有助于在保证防洪安全的前提下充分利用洪水退水期的水资源补充坝内库容,对提高汛期洪水利用效率具有重要意义。研究考虑社会目标、生态目标和效率目标,建立梯级橡胶坝群多目标蓄洪调度模型。[方法]为提升模型求解性能,提出了一种混合多目标粒子群算法(HMOPSO)用于模型求解,该算法通过Logistic映射初始化种群,采用差分进化策略优化迭代过程,并引入轮盘赌法选择全局最优解。以阜新市细河城市中心段梯级橡胶坝实际工程为例,选取该区域1994年和2013年的典型洪水过程为条件进行多目标蓄洪调度模型求解,将HMOPSO算法结果与MOPSO和NSGA-Ⅱ算法结果进行对比,利用多种性能指标评价各算法的Pareto前沿,并对调度解集进行规律性分析。[结果]各算法均满足求解需求,而HMOPSO算法相较于其他算法的GD指标优越28.57%以上,HV指标优越19.96%以上,证明HMOPSO算法在收敛性、均匀性和多样性方面均优于其他对比算法,能更有效应对不同洪水情景下的蓄洪调度需求;生态目标与效率目标之间存在负相关关系,若侧重于生态目标的实现,则整体调度时间会增加;洪水退水期的线形特征对调度的解集范围具有影响,洪尾可蓄水量、可调度时间越长,可行解集范围则越大。[结论]基于HMOPSO算法为城市段河道梯级橡胶坝群的合理蓄洪调度提供了理论依据,可为类似地区的研究提供参考。[Objective]The flood storage and scheduling problem of a cascade rubber dam group in urban river channels has nonlinearity,multidimensionality,and high constraints.Studying on its efficient and stable solution method can help fully utilize the water resources during the flood retreat period to supplement the reservoir capacity in the dam while ensuring flood control safety.It is of great significance to improve the efficiency of flood water utilization during the flood season.The study considered social,ecological,efficiency objectives and established a multi-objective flood storage and scheduling model for a cascade rubber dam group.[Methods]To improve the performance of model solving,a hybrid multi-objective particle swarm algorithm(HMOPSO)which initializes the population through logistic mapping is proposed,it can adopt differential evolution strategy to optimize the iterative process and introduce the roulette wheel method to select the global optimal solution.Taking the actual project of the cascade rubber dam in the central section of Xihe,Fuxin City as an example,the typical flood processes from 1994 to 2013 were selected as conditions to solve the multi-objective flood storage scheduling model.The results of HMOPSO algorithm were compared with MOPSO and NSGA-II algorithms,and various performance indicators were used to evaluate the Pareto front of each algorithm.The scheduling solution set was analyzed for regularity.[Results]The results showed that all algorithms meet the solving requirements,and the HMOPSO algorithm had a GD index superior to other algorithms by more than 28.57%and an HV index superior to other algorithms by more than 19.96%.This proves that the HMOPSO algorithm is superior to other comparative algorithms in terms of convergence,uniformity,and diversity,and can more effectively respond to flood storage and scheduling needs in different flood scenarios;there is a negative correlation between ecological goal and efficiency goal.If we focus on achieving ecological goal,the overall scheduli

关 键 词：梯级橡胶坝群 蓄洪调度模型 多目标粒子群算法 洪水资源利用 

分 类 号：TV213.9[水利工程—水文学及水资源]