A novel meta-heuristic optimization algorithm based on cell division: cell division optimizer

Optimization algorithms are crucial for solving some of the most intricate problems in engineering and science. Among these methods, meta-heuristic methods use stochastic elements to reach convergence, which is helpful in problems where the landscape is too complex for deterministic exploration. Thi...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Soft computing (Berlin, Germany) Ročník 29; číslo 13-14; s. 4879 - 4913
Hlavní autori: Jain, Sehej, Bharti, Kusum Kumari
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Berlin/Heidelberg Springer Berlin Heidelberg 01.07.2025
Springer Nature B.V
Predmet:
Algorithms
Artificial Intelligence
Benchmarks
Cell division
Computational Intelligence
Control
Convergence
Decision making
Engineering
Evolution
Feedback
Foraging behavior
Genetic algorithms
Heuristic
Heuristic methods
Mathematical Logic and Foundations
Mechatronics
Mitosis
Mutation
Optimization
Optimization algorithms
Physics
Robotics
Social behavior
Statistical analysis
Optimization techniques
Combinatorial optimization
Cell division optimizer
Engineering optimization
Meta-heuristics
ISSN:1432-7643, 1433-7479
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Optimization algorithms are crucial for solving some of the most intricate problems in engineering and science. Among these methods, meta-heuristic methods use stochastic elements to reach convergence, which is helpful in problems where the landscape is too complex for deterministic exploration. This paper introduces Cell Division Optimizer (CDO), a new meta-heuristic algorithm inspired by the two cell division processes: mitosis and meiosis. Mitosis is used to model exploitation whereas Meiosis is used for exploration. The method incorporates genetic diversity using mutation and crossovers, enhancing the algorithm’s ability to avoid premature convergence at local optima. The proposed method is evaluated over 50 well-known benchmark functions which include simple, unimodal, multimodal, and complex landscape benchmarks. Experiments are also performed for two classical engineering problems to verify the applications of CDO in real-world problems. This performance is compared with eight state-of-the-art algorithms including recently proposed methods and their specialized variants. A statistical analysis was performed to verify the significance of the results. Statistical analysis of 390 sub-experiments shows that CDO significantly outperformed competitors in 280 cases. In our experiments, it is seen that CDO consistently results in giving better solutions which is verified by studying the standard deviation among the results of the experiments. Finally, a convergence analysis is conducted for all the chosen methods. In most experiments, CDO converges to the global best earlier than the other algorithms, while avoiding local optima values.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1432-7643
1433-7479
DOI:10.1007/s00500-025-10670-4