A survey of recently developed metaheuristics and their comparative analysis

The aim of this study was to gather, discuss, and compare recently developed metaheuristics to understand the pace of development in the field of metaheuristics and make some recommendations for the research community and practitioners. By thoroughly and comprehensively searching the literature and...

Full description

Saved in:
Bibliographic Details
Published in:Engineering applications of artificial intelligence Vol. 117; p. 105622
Main Author: Alorf, Abdulaziz
Format: Journal Article
Language:English
Published: Elsevier Ltd 01.01.2023
Subjects:
Engineering optimization
Marine predator algorithm
Metaheuristics survey
Optimization algorithm comparison
Optimization algorithms
Optimization survey
Political optimizer
Political optimizer
Engineering optimization
Marine predator algorithm
Metaheuristics survey
Optimization algorithms
Optimization survey
Optimization algorithm comparison
ISSN:0952-1976, 1873-6769
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The aim of this study was to gather, discuss, and compare recently developed metaheuristics to understand the pace of development in the field of metaheuristics and make some recommendations for the research community and practitioners. By thoroughly and comprehensively searching the literature and narrowing the search results, we created with a list of 57 novel metaheuristic algorithms. Based on the availability of the source code, we reviewed and analysed the optimization capability of 26 of these algorithms through a series of experiments. We also evaluated the exploitation and exploration capabilities of these metaheuristics by using 50 unimodal functions and 50 multimodal functions, respectively. In addition, we assessed the capability of these algorithms to balance exploration and exploitation by using 29 shifted, rotated, composite, and hybrid CEC-BC-2017 benchmark functions. Moreover, we evaluated the applicability of these metaheuristics on four real-world constrained engineering optimization problems. To rank the algorithms, we performed a nonparametric statistical test, the Friedman mean rank test. Based on the statistical results for the unimodal and multimodal functions, we declared that the GBO, PO, and MRFO algorithms have better exploration and exploitation capabilities. Based on the results for the CEC-BC-2017 benchmark functions, we found the MPA, FBI, and HBO algorithms to be the most balanced. Finally, based on the results for the constrained engineering optimization problems, we declared that the HBO, GBO, and MA algorithms are the most suitable. Collectively, we confidently recommend the GBO, MPA, PO, and HBO algorithms for real-world optimization problems.
ISSN:0952-1976
1873-6769
DOI:10.1016/j.engappai.2022.105622