Personal Best Oriented Constriction Type Particle Swarm Optimization

In this paper, a new search strategy for constriction type particle swarm optimization is presented. The modification is based on the observation that personal past best experience is helpful for searching optimal result. As a result, instead of moving particle to the vicinity of current position, by letting the particle to explore the proximity of personal best position, a great improvement in computation efficiency and quality is achieved. The results are verified through testing on benchmark functions. The advantage of this new scheme is that no extra mathematic operation is introduced compared to those modifications proposed in literature