Particle Swarm Optimization with feasibility rules in constrained numerical optimization. A brief review

Abstract

Particle swarm optimization (PSO) is a population-based stochastic algorithm. The social behavior of a bird flock is the main inspiration of PSO. It was originally introduced to solve unconstrained optimization problems. However, due to the demands of real-world problems, PSO has evolved to be applied in constrained numerical optimization problems (CNOPs). Considering the fact that the set of feasibility rules is one of the most popular techniques to cope with constrains, it has been used extensively in those PSO-based algorithms. This paper presents a literature review of PSO for CNOPs in which, the conclusions suggest that the original PSO has changed to avoid its own disadvantages as premature convergence and also that some methods related to the inertia weight, constriction factor, additional operators, or hybridization with other metahuristics have been applied to improve the results in complex problems.