Two-Layers Particle Swarm Optimizer

Particle Swarm Optimization (PSO) has been recognized as one of the most widely used algorithms. One limitation of PSO is that it suffers from the problem of premature convergence which makes it easily trapped into local optima regions. To enhance its performances, this work presents a new PSO-based two layers optimizer. One layer is dedicated for performing the task of global search and the second layer is used for local search. Each layer consists of three search operations which are exploration, jump out, and convergence. To assess the performances of the proposed optimizer, CEC2010 large-scale optimization benchmark problems were used in this study. The outcomes indicated that the proposed optimizer outperformed other PSO variant algorithms in both convergence rate and average fitness value.