A niche-based evolutionary algorithm with dual cooperative archive for solving constrained multi-objective optimization problems

doi:10.1016/j.eij.2023.100422

Return

A niche-based evolutionary algorithm with dual cooperative archive for solving constrained multi-objective optimization problems

2024-03-01

Fengyu Guo

李和成 (Hecheng Li) *

DOI：10.1016/j.eij.2023.100422

Original

Save

View PDF

Abstract

En 中文

Constrained multi-objective optimization problems (CMOPs) are commonly encountered in engineering practice. The key to effectively solving these problems lies in achieving a timely balance between convergence, diversity, and feasibility during iterations. Furthermore, the appropriate utilization of infeasible solutions is crucial for identifying potential feasible regions. In order to accomplish this comprehensive objective, we propose a novel dual-stage constrained multi-objective evolutionary algorithm (CMOEA) called NACMOEA in this paper. It can be characterized by the following features: 1) Introducing a novel niche-based individual selection and infeasible solution utilization strategy to enhance convergence, diversity, and feasibility. 2) Presenting a cooperative search strategy assisted by dual archives to approximate the constrained Pareto front (CPF) from both feasible and infeasible perspectives, thereby improving the efficiency of obtaining the complete CPF. 3) Designing a new stage switch method based on non-dominant coverage rate to ensure proper completion of search stage switching. Extensive experiments demonstrate that NACMOEA exhibits competitive comprehensive performance when compared with other advanced CMOEAs.

Keywords:

Constrained multi-objective optimization

Dual stage algorithm

Infeasible individual utilization

Niche-based selection method

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.