A Surrogate-Assisted Two-Stage Differential Evolution for Expensive Constrained Optimization
A Surrogate-Assisted Two-Stage Differential Evolution for Expensive Constrained Optimization
复制标题
用于昂贵约束优化的代理辅助两阶段差分进化
DOI：
10.1109/tetci.2023.3240221
复制
发表时间：
2023-06
期刊：
影响因子：
--
通讯作者：
Tianzi Zheng
中科院分区：
文献类型：
--
作者：
Yuanchao Liu;Jianchang Liu;Yaochu Jin;Fei Li;Tianzi Zheng
代理辅助进化算法（SAEAs）已成功应用于代价高昂的优化问题。然而，大多数代理辅助进化算法是为代价高昂的无约束优化而设计的，对于带有不等式（inequ后面似乎不完整，可能是inequality即不等式相关内容）的代价高昂的优化问题关注较少。
Surrogate-assisted evolutionary algorithms (SAEAs) have been successfully employed for expensive optimization. However, most SAEAs are designed for expensive unconstrained optimization, and less attention has been paid to expensive optimization with inequality constraints. Therefore, this work proposes a novel SAEA, called surrogate-assisted two-stage differential evolution (SA-TSDE), for expensive constrained optimization. In the first search stage, surrogate-assisted hybrid differential evolution is adopted for prescreening promising solutions in the decision space for exploration. Moreover, an effective repair strategy, named surrogate based repair strategy, is introduced to move the infeasible solutions closer to the feasible region. In the second search stage, a clustering strategy of feasible solutions is proposed based on the information provided by the first search stage and historical search. The clustering strategy adaptively generates a number of clusters, each of which can form a promising local region for the local search. Afterwards, local surrogates are built for finding the predicted optima in each local region. During the search process, a good balance between exploration and exploitation can be obtained by interleaving the two search stages. Experimental results indicate that SA-TSDE is highly competitive compared with some state-of-the-art methods.
登录
查看更多内容
影响因子：
5.7
作者：
Grimaccia, Francesco;Mussetta, Marco;Zich, Riccardo E.
通讯作者：
Zich, Riccardo E.
DOI：
10.1016/j.cosrev.2009.07.001
发表时间：
2009-08
期刊：
Comput. Sci. Rev.
影响因子：
--
作者：
S. Salcedo-Sanz
通讯作者：
S. Salcedo-Sanz
DOI：
10.1145/1830761.1830809
发表时间：
2010-07
期刊：
--
影响因子：
--
作者：
Amy FitzGerald;D. O'Donoghue
通讯作者：
Amy FitzGerald;D. O'Donoghue
DOI：
10.1016/j.eswa.2022.119075
发表时间：
2022-10
期刊：
Expert Syst. Appl.
影响因子：
--
作者：
Yuanchao Liu;Jianchang Liu;Shubin Tan
通讯作者：
Yuanchao Liu;Jianchang Liu;Shubin Tan
影响因子：
3.9
作者：
Anirban Basudhar;C. Dribusch;Sylvain Lacaze;S. Missoum
通讯作者：
Anirban Basudhar;C. Dribusch;Sylvain Lacaze;S. Missoum