Scalars Impact on Particle Swarm Optimization Performance

The Particle swarm optimization (PSO) is optimization technique that incorporates swarming behaviors observed in insects, birds and fish. PSO optimizes an objective function by undertaking a population-based search. In this technique only few parameter are need to be tuned. The work on PSO considers inertia weight, two constant multiplier terms known as ¿self confidence¿ and ¿swarm confidence¿, maximum velocity Vmax and the swarm size as the main parameters. This paper analyzes the impact of scalars r1 and r2 on the performance of the particle swarm optimizer. Self confidence, swarm confidence and the maximum velocity which restricts Velocity of each particle within the [-Vmax, Vmax] interval are kept constant for all iteration by varying the inertia weight in each iteration. The scalars r1 and r2 have positive impact on the performance of PSO.