Swarm-Genetics: A Hybrid PSO-GA Regeneration Model for Global Optimization Benchmark Problems

Abstract

Particle Swarm Optimization (PSO) is a highly effective algorithm for solving complex optimization problems in multidimensional spaces using particle-based and population-based approaches. However, a common limitation of PSO is its tendency to get trapped in local optima. This study proposes SWARM-GENETICS, an adaptive flock regeneration framework that integrates Genetic Algorithm (GA) into the PSO process to overcome this limitation. By simulating an evolutionary regeneration mechanism within the swarm, the framework enhances diversity and exploration capabilities in the solution space. The GA component supports the regeneration of stagnated particles, improving convergence efficiency and preventing premature stagnation. This approach is tested on several benchmark optimization functions characterized by multiple local optima. The results demonstrate that the proposed hybrid model significantly improves performance in escaping local traps and exploring deeper regions of the solution space, leading to more optimal outcomes.