Switching Constraint Handling Evolutionary Algorithm for Constrained Multi-Modal Multi-Objective Optimization Problems

In real-world applications, constrained multi-modal multi-objective optimization problems (CMMOPs) are prevalent across various science and engineering domains. Multi-objective evolutionary algorithms (MOEAs) are well-suited for solving such challenging problems, owing to their versatility and effic...

Full description

Saved in:
Bibliographic Details
Published in2024 SICE Festival with Annual Conference (SICE FES) pp. 1079 - 1084
Main Authors Ono, Yuhiro, Harada, Tomohiro, Miura, Yukiya
Format Conference Proceeding
LanguageEnglish
Published The Society of Instrument and Control Engineers - SICE 27.08.2024
Subjects
Benchmark testing
Constraint handling
EA CMMOP SPEA2 CMMOCEA SCHMMEA
Evolutionary computation
Iron
Optimization
Search problems
Switches
Online AccessGet full text

Cover

More Information
Summary:In real-world applications, constrained multi-modal multi-objective optimization problems (CMMOPs) are prevalent across various science and engineering domains. Multi-objective evolutionary algorithms (MOEAs) are well-suited for solving such challenging problems, owing to their versatility and efficient search capabilities. This paper proposes the switching constraint handling multi-modal multi-objective evolutionary algorithm (SCHMMEA), a MOEA for CMMOPs that switches the constraint-handling strategies during the search process. Specifically, SCHMMEA initially explores high-quality solutions without considering constraints, then the late stage considers the constraints. We compare the proposed method with CMMOCEA, a conventional method, on 14 CMMOP benchmarks. In addition, we examine different parameters switching the constraint handling strategies in the proposed method. The experimental results show that SCHMMEA with the constraint handling at 2/3 of the maximum generations outperformed the conventional method by 13, 12, and 11 problems in IGDX, IGD, and HV, respectively.