Improvement of the particle swarm optimization algorithm has become increasingly important to deliver it out of local optima trapping and increase its convergence rate. In this paper a personal best adaptive weight is proposed as a new PSO variant named personal best adaptive weight particle swarm optimization (PBAW-PSO) to choose different inertia weight for different particles in the swarm to update their velocity. The proposed variant was compared with three other inertia weight improved variants on six benchmark functions. The comparison was done based on the best cost, mean cost, simulation time, standard deviation and convergence rate. The overall results showed that the PBAW-PSO variant had a better performance than the other variants.

Keywords : Metaheuristic; Inertia Weight; Evolutionary; Particle Swarm Optimization; Convergence.