Diversity Control in the Hybridization GA-PSO with Fuzzy Adaptive Inertial Weight
Rodrigo Possidônio Noronha · 2021
In this paper, a new stochastic optimization methodology based on hybridization involving Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) is proposed, with the goal of obtaining an optimization method with fast and non-premature convergence. The slow and premature convergence occurs due to inefficient control of position diversity in the search space. Diversity control can be achieved by parametric adaptation of the inertial weight. However, obtaining formulations that describe the dynamic behavior of the inertial weight in order to control diversity is not trivial. For this, in the proposed methodology, a Mamdami fuzzy system is used with the goal of adapting the inertial weight and thus to obtain a search process with a fast and non-premature convergence.