面向梯级水库多目标优化调度的进化算法研究  

Research on evolutionary algorithms for multi-objective optimal operation of cascade reservoirs

作　　者：纪昌明[1] 马皓宇 彭杨[1] JI Changming;MA Haoyu;PENG Yang(School of Water Resources and Hydropower Engineering,North China Electric Power University,Beijing 102206,China)

机构地区：[1]华北电力大学水利与水电工程学院,北京102206

出　　处：《水利学报》2020年第12期1441-1452,共12页Journal of Hydraulic Engineering

基　　金：国家重点研发计划项目(2016YFC0402309,2016YFC0402208)。

摘　　要：实际工程中以梯级水库多目标优化调度为代表的大规模高维多目标优化问题,其优化难度是一般方法所难以应对的。为此本文提出一种新型的多目标粒子群算法LMPSO,其包含了基于超体积指标Ihk的适应值分配方法与基于问题变换的搜索空间降维策略,以有效处理问题的高维目标向量与大规模决策变量。将该算法应用于溪洛渡-向家坝梯级水库的中长期多目标优化调度中,并与4种知名算法的计算结果进行对比分析,验证LMPSO在求解该类问题上的卓越性能。由此为多目标优化调度高质量非劣解集的获取提供一种可靠的方法,并为下一步的多目标调度决策提供有力的数据支持。In practical projects,the large-scale multi-objective optimization problems,represented by multi-objective optimal operation of cascade reservoirs,are difficult to be solved by general methods.Therefore,this paper proposes a new multi-objective particle swarm optimization(LMPSO),which includes both a fitness assignment approach based on the hypervolume indicator Ihkand a strategy to reduce search space dimensions based on problem transformation,in order to effectively deal with high-dimensional objective vectors and large-scale decision variables.The algorithm is applied to the medium-and-long-term multi-objective optimal operation of Xiluodu-Xiangjiaba cascade reservoirs,and the calculation results are comparatively analyzed with four established algorithms,to verify the excellent performance of LMPSO.It provides not only a reliable method to obtain high-quality Pareto solution sets in multi-objective optimal operation but also strong data support for the subsequent multi-objective operation decision-making.

关 键 词：多目标优化 梯级水库调度 大规模优化 高维多目标优化 

分 类 号：TV697.11[水利工程—水利水电工程]