摘要
针对求解昂贵超多目标黑箱优化问题的算法进行研究。基于SOCEMO算法,将ε-支配用于目标值采样策略,提出ε-SOCEMO算法。在42个超多目标测试问题上对ε-SOCEMO算法和SOCEMO算法进行了数值实验,结果显示ε-SOCEMO算法在性能评价指标上具有较好的表现。数值实验表明这种改进的目标值采样策略可以提升SOCEMO算法在昂贵超多目标黑箱优化问题上的计算效果。
It investigates the algorithm for solving expensive many-objective black-box optimization problems.The algorithmε-SOCEMO is proposed based on the SOCEMO algorithm,which usesε-dominance for the target-value sampling strategy.Numerical experiments on algorithmsε-SOCEMO and SOCEMO were performed on 42 many-objective test problems,and the results show thatε-SOCEMO has better performance in the metrics.Numerical experiments show that this modified target-value sampling strategy can improve the computational effectiveness of the SOCEMO algorithm on expensive many-objective black-box optimization problems.
作者
魏玉涛
白富生
WEI Yutao;BAI Fusheng(School of Mathematical Sciences,Chongqing Normal University;National Center for Applied Mathematics,Chongqing Normal University,Chongqing 40133l,China)
出处
《重庆师范大学学报（自然科学版）》
CAS
北大核心
2023年第4期15-22,共8页
Journal of Chongqing Normal University:Natural Science
基金
重庆市教育委员会科学技术研究计划重点项目(No.KJZD-K202114801)
重庆市技术创新与应用发展专项(No.cstc2021jscx-jbgsX0001)
重庆市自然科学基金创新发展联合基金项目(No.2022NSCQ-LZX0301)。
关键词
多目标优化
昂贵黑箱函数
响应面方法
径向基函数
many-objective optimization
expensive black box functions
surrogate methods
radial basis functions
