Multi-Population Optimization Framework Based on Plant Evolutionary Strategy and Its Application to Engineering Design Problems

Optimization problems are widespread across various fields, including industry, agriculture, and healthcare. Metaheuristic algorithms (MAs) are commonly employed to solve these problems due to their flexibility and robustness. However, despite their success, MAs inspired by plant evolutionary strate...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:International journal of computational intelligence systems Jg. 18; H. 1; S. 117 - 22
Hauptverfasser: Cheng, Hongwei, Li, Jun, Zhang, Xiaoming, Li, Tingjuan, Zhang, Panpan
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Dordrecht Springer Netherlands 15.05.2025
Springer Nature B.V
Springer
Schlagworte:
Adaptation
Algorithms
Artificial Intelligence
Computational Intelligence
Control
Convergence
Design
Design engineering
Design optimization
Engineering
Environmental conditions
Evolution
Exploitation
Global optimization
Heuristic methods
Mathematical Logic and Foundations
Mechatronics
Metaheuristic algorithms
Multi-population evolution
Optimization algorithm framework
Optimization algorithms
Performance evaluation
Plant population evolution
Plant populations
Research Article
Robotics
Robustness (mathematics)
Seeds
Strategy
Task complexity
Multi-population evolution
Optimization algorithm framework
Plant population evolution
Metaheuristic algorithms
ISSN:1875-6883, 1875-6891, 1875-6883
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Optimization problems are widespread across various fields, including industry, agriculture, and healthcare. Metaheuristic algorithms (MAs) are commonly employed to solve these problems due to their flexibility and robustness. However, despite their success, MAs inspired by plant evolutionary strategies remain underexplored. This paper introduces a novel multi-population optimization framework based on the plant evolutionary strategy (PES_MPOF), which leverages plant evolutionary principles to improve optimization performance by maintaining population diversity and accelerating convergence in complex tasks. PES_MPOF integrates multiple subpopulations, each evolving according to different plant evolutionary models. These subpopulations mimic natural distribution and reproduction strategies, fostering solution diversity through both cooperation and competition. Additionally, PES_MPOF adapts population parameters based on the evolutionary performance of subpopulations, further enhancing its robustness and efficiency. The PES_MPOF algorithm was tested on the IEEE CEC 2020 benchmark suite and several classic engineering design problems. It outperforms other state-of-the-art optimization algorithms, demonstrating significant improvements in global optimization, solution accuracy, and convergence speed. PES_MPOF effectively addresses the challenges of premature convergence and loss of diversity, making it a robust and efficient optimization tool. Its innovative multi-population framework, inspired by plant evolutionary strategies, enhances both exploration and exploitation. Experimental results validate its effectiveness across a broad range of optimization problems, including those with constraints. The part of algorithm’s code will be made available upon the paper’s acceptance: https://github.com/ChengHongwei430/PES_MPOF .
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1875-6883
1875-6891
1875-6883
DOI:10.1007/s44196-025-00779-7