Hybridized Particle Swarm Optimization for Constrained Problems

This paper presents hybridized implementation of the well-known particle swarm optimization algorithm that belongs to the family of swarm intelligence metaheuristics. The proposed approach was adapted for tackling constrained optimization problems. With the basic goals to enhance the converge of the algorithm and to improve the exploitation – exploration tradeoff, the mechanism that replaces exhausted solutions from the population with randomly generated solutions from the search domain was adopted from the artificial bee colony approach. Proposed metaheuristic was tested on standard constrained engineering benchmark, and comparative analysis with other state-of-the-art algorithms was conducted. Empirical results obtained from practical simulations proved that the hybridized particle swarm optimization for constrained problems is able to successfully tackle this type of NP hard challenges.