MODIFICATIONS OF THE PARTICLE SWARM OPTIMIZATION AND NEW PROPOSED VARIANT

Particle swarm optimization (PSO) is a stochastic meta-heuristic computational technique which serves to find the best region of a multidimensional space. The method belongs to the group of evolutionary computation, to the subgroup of swarm intelligence. It is based on an iterative work with a population and mimics the movement of school of fish or flock of birds. The low number of parameters to adjust and easy implementation are the main benefits of the PSO method. The main goal of this paper is to provide a comprehensive review about different variants of the particle swarm optimization. The original PSO algorithm is modified due to improve the optimization ability and to avoid the premature convergence to the local minimum. The parameter of inertia weight or constriction factor was implemented to the equation. In this paper, new modification of PSO algorithm is presented, where the velocity of individuals depends on the nearest neighbourhood of each particle. The comparison between the proposed algorithm and other existing methods is presented and discussed in the paper.