A Case Study on Hybridizing Differential Evolution Variants in The Field of Optimization Using Benchmark Functions

This paper aims to improve optimization efficiency by investigating the hybridization of Differential Evolution (DE) variants using an Island Model framework. DE, a population-based meta-heuristic algorithm, is widely recognized for its simplicity and robustness in solving complex optimization problems. However, traditional DE variants often struggle with premature convergence and the balance between exploration and exploitation, limiting their effectiveness in multimodal landscapes. To address these challenges, this study introduces hybrid algorithms that integrate multiple DE strategies within the Island Model framework. The primary objective is to enhance solution diversity and convergence reliability by enabling independent evolution of sub-populations (islands) with periodic migrations to exchange information. This cooperative approach prevents premature convergence and achieves a better balance between exploration and exploitation. The performance of the proposed hybrid DE algorithms is rigorously evaluated on a set of standard benchmark functions, demonstrating significant improvements in convergence speed and solution accuracy compared to traditional DE approaches. These findings underscore the potential of the Island Model and adaptive hybridization techniques in advancing optimization methods for complex problem landscapes.