文章摘要
艾学山,郭佳俊,穆振宇,陈森林,杨百银,周鹏程.梯级水库群多目标优化调度模型及CPF-DPSA算法研究[J].水利学报,2023,54(1):68-78
梯级水库群多目标优化调度模型及CPF-DPSA算法研究
Research on multi-objective optimal operation model of cascaded hydropower system and CPF-DPSA algorithm
投稿时间：2021-10-21  
DOI：10.13243/j.cnki.slxb.20210947
中文关键词: 梯级水库群  多目标  CPF-DPSA  惩罚因子  非劣解集
英文关键词: multi-objective  cascaded hydropower system  CPF-DPSA  penalty factors  non-dominated solution set
基金项目:国家自然科学基金项目(51779177);水电水利规划设计总院项目(KJHT-SD2019-08-04)
作者单位
艾学山 武汉大学 水资源与水电工程科学国家重点实验室, 湖北 武汉 430072
海绵城市建设水系统科学湖北省重点实验室(武汉大学), 湖北 武汉 430072 
郭佳俊 武汉大学 水资源与水电工程科学国家重点实验室, 湖北 武汉 430072 
穆振宇 武汉大学 水资源与水电工程科学国家重点实验室, 湖北 武汉 430072 
陈森林 武汉大学 水资源与水电工程科学国家重点实验室, 湖北 武汉 430072 
杨百银 水电水利规划设计总院, 北京 100120 
周鹏程 中国电建集团昆明勘测设计研究院有限公司, 云南 昆明 650051 
摘要点击次数: 1484
全文下载次数: 930
中文摘要:
      梯级水库群多目标联合优化调度是水能资源高效利用的重要研究内容,现有水库群多目标优化调度模型及求解算法的通用性亟待提高。在总结现有三种目标函数型式(累积值、极值及百分比)的基础上,以梯级总发电量最大、最小出力最大和生态断面用水保证率最大为目标,建立了具有普适性的梯级水库群多目标优化调度模型,提出了求解该模型的基于惩罚因子的动态规划逐次逼近算法(CPF-DPSA),探究了各目标与对应惩罚系数之间的变化关系,确定了各惩罚系数的影响范围,获得了分布较为均匀和广泛的非劣解集。老挝南欧江梯级水库群应用表明,该模型具有较好的适用性,CPF-DPSA算法获得的非劣解集分布广泛、均匀。尤其在长系列优化方面,CPF-DPSA算法在结果精度、非劣解质量和非劣解分布等方面比第三代快速非劣排序遗传算法(NSGA-Ⅲ)表现出更好的性能。
英文摘要:
      The multi-objective optimal operation of cascaded hydropower system is an important research content in the efficient utilization of hydropower resources, in which the universality of the existing multiobjective optimal operation model and its solving algorithm is urgent to improve. On the basis of summarizing the existing three types of objective function (cumulative value, extreme value and percentage), the multi-objective optimal operation model of cascade hydropower system with universal applicability is established with the maximum of power generation, the maximum of the minimum output power and the maximum of water supply guarantee rate of ecological section, and the dynamic programming successive approximation method based on the combination of penalty factors (CPF-DPSA) is proposed to solve the model. This algorithm explores the relationship between each target and the corresponding factor, confirms the scope of influence of each penalty coefficient, and obtains the non-dominated solution set with relatively uniform distribution. The application of Nam Ou cascaded hydropower system in Laos shows that this model has good applicability and the non-dominated solution set obtained by CPF-DPSA algorithm is distributed widely and uniformly. Especially in the long series problem, the algorithm has better performance than the third generation of fast non-dominated sorting genetic algorithm (NSGA-Ⅲ) in terms of result accuracy, non-inferior solution quality and Pareto front distribution.
查看全文   查看/发表评论  下载PDF阅读器
关闭