A Multi-strategy Improved Snake Optimizer Assisted with Population Crowding Analysis for Engineering Design Problems

Snake Optimizer (SO) is a novel Meta-heuristic Algorithm (MA) inspired by the mating behaviour of snakes, which has achieved success in global numerical optimization problems and practical engineering applications. However, it also has certain drawbacks for the exploration stage and the egg hatch pr...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of bionics engineering 2024-05, Vol.21 (3), p.1567-1591
Hauptverfasser: Peng, Lei, Yuan, Zhuoming, Dai, Guangming, Wang, Maocai, Li, Jian, Song, Zhiming, Chen, Xiaoyu
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Snake Optimizer (SO) is a novel Meta-heuristic Algorithm (MA) inspired by the mating behaviour of snakes, which has achieved success in global numerical optimization problems and practical engineering applications. However, it also has certain drawbacks for the exploration stage and the egg hatch process, resulting in slow convergence speed and inferior solution quality. To address the above issues, a novel multi-strategy improved SO (MISO) with the assistance of population crowding analysis is proposed in this article. In the algorithm, a novel multi-strategy operator is designed for the exploration stage, which not only focuses on using the information of better performing individuals to improve the quality of solution, but also focuses on maintaining population diversity. To boost the efficiency of the egg hatch process, the multi-strategy egg hatch process is proposed to regenerate individuals according to the results of the population crowding analysis. In addition, a local search method is employed to further enhance the convergence speed and the local search capability. MISO is first compared with three sets of algorithms in the CEC2020 benchmark functions, including SO with its two recently discussed variants, ten advanced MAs, and six powerful CEC competition algorithms. The performance of MISO is then verified on five practical engineering design problems. The experimental results show that MISO provides a promising performance for the above optimization cases in terms of convergence speed and solution quality.
ISSN:1672-6529
2543-2141
DOI:10.1007/s42235-024-00505-7