.png){kind=icon}
返回
A surrogate-assisted expensive constrained multi-objective global optimization algorithm and application
.png){kind=icon}
PRE
AI
DOI：10.1016/j.asoc.2024.112226.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Expensive multi-objective optimization problems (MOPs) have seen the successful applications of surrogateassisted evolutionary algorithms (SAEAs). Nevertheless, the majority of SAEAs are developed for costly unconstrained optimization, and costly constrained MOPs (CMOPs) have received less attention. Therefore, this article proposes a surrogate-assisted global optimization algorithm (named CTEA) for solving CMOPs within a very limited number of fitness evaluations. The proposed algorithm combines two selection frameworks, a bi-level selection framework, and an adaptive sampling framework, to enhance optimization performance. Leveraging on a constraint-improving strategy and a Pareto-based three-indicator criterion (convergence, constraint, and diversity indicators) at the different levels, the proposed bi-level selection framework can select more promising solutions. Moreover, an adaptive sampling framework is developed to prioritize objective and constraint functions and select the candidate solutions for real function evaluations according to the priority. Experimental results demonstrate that the proposed CTEA exhibits superior performance when compared with five state-of-theart algorithms, achieving the best results in 61.9% out of the 64 test instances. Finally, CTEA is applied to the multidisciplinary design optimization of blended-wing-body underwater gliders, and an impressive solution set is obtained.
Keyword:
Bi-level selection
Adaptive sampling
Global optimization
Expensive constrained multi-objective
Blended-wing-body underwater glider
期刊
IF:
6.6
论文数:
1.4W
被引数:
4.8W
机构
引用论文
.png){kind=icon}
PlatEMO: A MATLAB Platform for Evolutionary Multi-Objective OptimizationPlatEMO: 用于进化多目标优化的MATLAB平台
OAAI
OAAI分享收藏Multiobjective evolutionary algorithms: A comparative case study and the Strength Pareto approach多目标进化算法: 比较案例研究和强度帕累托方法
PREAI
分享收藏
查看更多内容