Dragonfly algorithm: a comprehensive review and applications
- PDF / 1,664,612 Bytes
- 22 Pages / 595.276 x 790.866 pts Page_size
- 1 Downloads / 234 Views
(0123456789().,-volV) (0123456789().,-volV)
REVIEW
Dragonfly algorithm: a comprehensive review and applications Yassine Meraihi1 • Amar Ramdane-Cherif2 • Dalila Acheli3 • Mohammed Mahseur4 Received: 21 September 2019 / Accepted: 14 March 2020 Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract Dragonfly algorithm (DA) is a novel swarm intelligence meta-heuristic optimization algorithm inspired by the dynamic and static swarming behaviors of artificial dragonflies in nature. It has proved its effectiveness and superiority compared to several well-known meta-heuristics available in the literature. This paper presents a comprehensive review of DA and its new variants classified into modified and hybrid versions. It also describes the main diverse applications of DA in several fields and areas such as machine learning, neural network, image processing, robotics, and engineering. Finally, the paper suggests some possible interesting research on the applications and hybridizations of DA for future works. Keywords Dragonfly algorithm Nature-inspired algorithm Swarm intelligence Meta-heuristics
1 Introduction Meta-heuristics are approximate optimization algorithms that have attracted great interest from many researchers in several fields such as computer science, operation research, bio-informatics, and engineering. This interest is due to their simplicity, flexibility, and robustness to solve a variety of optimization problems in a reasonable time.
& Yassine Meraihi [email protected] Amar Ramdane-Cherif [email protected] Dalila Acheli [email protected] Mohammed Mahseur [email protected] 1
LIST Laboratory, University of M’Hamed Bougara Boumerdes, Avenue of Independence, 35000 Boumerdes, Algeria
2
LISV Laboratory, University of Versailles St-Quentin-enYvelines, 10-12 Avenue of Europe, 78140 Velizy, France
3
LAA Laboratory, University of M’Hamed Bougara Boumerdes, Avenue of Independence, 35000 Boumerdes, Algeria
4
Faculty of Electronics and Informatics, University of Sciences and Technology Houari Boumediene, El Alia Bab Ezzouar, 16025 Algiers, Algeria
According to Fister et al. [1], meta-heuristics can be divided into two categories: non-nature-inspired metaheuristics and nature-inspired meta-heuristics. Nature-inspired meta-heuristic algorithms can be classified into five main categories: evolutionary-based, physics-based, chemistry-based, human-based, and swarm intelligence-based. Evolutionary algorithms are inspired by the concept of biological evolution in nature using the operators of selection, crossover, mutation, and reproduction to find better candidate solutions. Genetic algorithm (GA) [2], developed by Holland in 1992, is regarded as the best evolutionary algorithm. Some other popular evolutionary algorithms are: differential evolution (DE) [3], evolutionary programming (EP) [4], evolution strategy (ES) [5], genetic programming (GP) [6], probability-based incremental learning (PBIL) [7], and biogeography-based optimizer (BBO) [8, 9]. The second catego
Data Loading...
