A Survey on Particle Swarm Optimization Algorithm

Abstract

Particle swarm optimization (PSO) is a computational method for finding optimal solutions in a random search space. PSO is a heuristic global optimization method, suggested by Kennedy with Eberhart in the year of 1995. The suggested swarm algorithm (PSO) is inspired from the behavior of swarms living in self-organized groups which fly to search for food sources. In the present time, PSO approach is the most commonly used optimization technique due to its simplicity and efficient performance. The mathematical model of the algorithm uses the velocity, position and fitness of each solution. By using the velocity equation, the group of particles moves in search space. Fitness evaluation function evaluates optimality of the solution. For this reason, researchers are continuously improving the capability of velocity and position counting methods to improve the performance of PSO algorithms. There are lots of methods to navigate the algorithm in the proper search direction by tuning position and velocity equation parameters. This survey paper presents a comprehensive and systematic study on the most basic as well as some of the very recent variants of particle swarm optimization algorithms like classical PSO (CPSO), PSO with time varying acceleration coefficient (PSO-TVAC) and many more approaches. The paper includes a tabular comparison of the various PSO variants and this would be beneficial to the researchers, who are working in the field of swarm intelligence.