New particle swarm optimizer with sigmoid increasing inertia weight

Abstract

The inertia weight of particle swarm optimization (PSO) is a mechanism to control the exploration and exploitation abilities of the swarm and as mechanism to eliminate the need for velocity clamping. The present paper proposes a new PSO optimizer with sigmoid increasing inertia weight. Four standard non-linear benchmark functions are used to confirm its validity. The comparison has been simulated with sigmoid decreasing and linearly increasing inertia weight. From experiments, it shows that PSO with increasing inertia weight gives better performance with quick convergence capability and aggressive movement narrowing towards the solution region.