Modified Archimedes optimization algorithm for global optimization problems: a comparative study

Basit Öğe Kaydı
dc.contributor.authorNurmuhammed, Mustafa
dc.contributor.authorAkdag, Ozan
dc.contributor.authorKaradag, Teoman
dc.date.accessioned2026-06-19T06:41:15Z
dc.date.available2026-06-19T06:41:15Z
dc.date.issued2024
dc.departmentMalatya Turgut Özal Üniversitesi
dc.description.abstractArchimedes Optimization Algorithm (AOA) is a recent optimization algorithm inspired by Archimedes' Principle. In this study, a Modified Archimedes Optimization Algorithm (MDAOA) is proposed. The goal of the modification is to avoid early convergence and improve balance between exploration and exploitation. Modification is implemented by a two phase mechanism: optimizing the candidate positions of objects using the dimension learning-based (DL) strategy and recalculating predetermined five parameters used in the original AOA. DL strategy along with problem specific parameters lead to improvements in the balance between exploration and exploitation. The performance of the proposed MDAOA algorithm is tested on 13 standard benchmark functions, 29 CEC 2017 benchmark functions, optimal placement of electric vehicle charging stations (EVCSs) on the IEEE-33 distribution system, and five real-life engineering problems. In addition, results of the proposed modified algorithm are compared with modern and competitive algorithms such as Honey Badger Algorithm, Sine Cosine Algorithm, Butterfly Optimization Algorithm, Particle Swarm Optimization Butterfly Optimization Algorithm, Golden Jackal Optimization, Whale Optimization Algorithm, Ant Lion Optimizer, Salp Swarm Algorithm, and Atomic Orbital Search. Experimental results suggest that MDAOA outperforms other algorithms in the majority of the cases with consistently low standard deviation values. MDAOA returned best results in all of 13 standard benchmarks, 26 of 29 CEC 2017 benchmarks (89.65%), optimal placement of EVCSs problem and all of five real-life engineering problems. Overall success rate is 45 out of 48 problems (93.75%). Results are statistically analyzed by Friedman test with Wilcoxon rank-sum as post hoc test for pairwise comparisons.
dc.description.sponsorshipInn niversitesi [FDK-2023-3163]; Inonu University-the Scientific Research Projects (BAP) Unit
dc.description.sponsorshipThis research is supported by Inonu University-the Scientific Research Projects (BAP) Unit (No. FDK-2023-3163). The Authors would like to thank Dr. Ahmet Kadir Arslan for comments and discussions which helped improving the quality of the paper.
dc.identifier.doi10.1007/s00521-024-09497-1
dc.identifier.issn0941-0643
dc.identifier.issn1433-3058
dc.identifier.orcid0000-0001-8163-8898
dc.identifier.orcid0000-0002-5957-3255
dc.identifier.orcid0000-0002-7682-7771
dc.identifier.scopus2-s2.0-85185968069
dc.identifier.scopusqualityQ1
dc.identifier.urihttps://doi.org/10.1007/s00521-024-09497-1
dc.identifier.urihttps://hdl.handle.net/20.500.12899/6135
dc.identifier.wosWOS:001170753100002
dc.identifier.wosqualityN/A
dc.indekslendigikaynakWeb of Science
dc.indekslendigikaynakScopus
dc.language.isoen
dc.publisherSpringer London Ltd
dc.relation.ispartofNeural Computing & Applications
dc.relation.publicationcategoryMakale - Uluslararası Hakemli Dergi - Kurum Öğretim Elemanı
dc.rightsinfo:eu-repo/semantics/openAccess
dc.snmzKA_WOS_20260612
dc.subjectBenchmark Functions
dc.subjectModified Optimization Algorithms
dc.subjectOptimization Algorithms
dc.subjectSwarm Intelligence
dc.titleModified Archimedes optimization algorithm for global optimization problems: a comparative study
dc.typeArticle

Dosyalar

Koleksiyon

WoS İndeksli Yayınlar Koleksiyonu
Scopus İndeksli Yayınlar Koleksiyonu