Emulous mechanism based multi-objective moth–flame optimization algorithm

In recent years, there has been growing interest in using metaheuristic algorithms to solve various complex engineering optimization problems. Most of the real-world problems comprise of more than one objective. Due to the inherent difficulty of such problems and lack of proficiency, researchers in...

Full description

Saved in:
Bibliographic Details
Published in:Journal of parallel and distributed computing Vol. 150; pp. 15 - 33
Main Authors: Sapre, Saunhita, S., Mini
Format: Journal Article
Language:English
Published: Elsevier Inc 01.04.2021
Subjects:
Constrained engineering design
Emulous learning
Moth–lame optimization
Multi-objective algorithm
Pareto-optimal solutions
Constrained engineering design
Multi-objective algorithm
Emulous learning
Pareto-optimal solutions
Moth–lame optimization
ISSN:0743-7315, 1096-0848
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In recent years, there has been growing interest in using metaheuristic algorithms to solve various complex engineering optimization problems. Most of the real-world problems comprise of more than one objective. Due to the inherent difficulty of such problems and lack of proficiency, researchers in different domains often aggregate multiple objectives and use single-objective optimization algorithms to solve them. However, the aggregation-based methods fail to solve the multi-objective problems (MOPs) effectively. Several multi-objective evolutionary algorithms (MOEAs) have been proposed and are being used to solve such problems in the past few years. In this paper, we propose an Emulous Mechanism-based multi-objective Moth–Flame Optimization (EMMFO) algorithm, where the moth positions are updated based on the pairwise competitions between the moths in each generation. The proposed EMMFO is tested on a diverse set of multi-objective benchmark functions like ZDT, DTLZ, WFG, CEC09 special session test suites and four constrained engineering design problems. The results are compared with various state-of-the-art multi-objective algorithms like NSGAII, SPEA2, PESA2, MOEA/D, MOPSO, MOACO, NSMFO, IEMO, CLPSO-LS, MOEA/D-CRA, PAL-SAPSO, and MORBABC/D. Extensive experimental results demonstrate superior optimization performance of the proposed algorithm.
ISSN:0743-7315
1096-0848
DOI:10.1016/j.jpdc.2020.12.010