A clustering-based surrogate-assisted evolutionary algorithm (CSMOEA) for expensive multi-objective optimization

This paper presents a novel surrogate-assisted evolutionary algorithm, CSMOEA, for multi-objective optimization problems (MOPs) with computationally expensive objectives. Considering most algorithms (SAEAs) do not make full use of population information and only in either the objective space or design independently, to address this limitation, we propose new strategy comprehensive utilization space. The proposed CSMOEA adopts an adaptive clustering divide current into good bad groups, centers are obtained, respectively. Then, bi-level sampling is select best samples both space, using distance approximated values radial basis functions. effectiveness compared five state-of-the-art on 21 widely used benchmark problems, results show high efficiency balance between convergence diversity. Additionally, applied shape blend-wing-body underwater gliders 14 decision variables two objectives, demonstrating its solving real-world engineering problems.