Incorporating the Avoidance Behavior to the Standard Particle Swarm Optimization 2011

Inspired from social and cognitive behaviors of animals living as swarms; particle swarm optimization (PSO) provides a simple but very powerful tool for researchers who are dealing with collective intelligence. The algorithm depends on modeling the very basic random behavior (i.e. exploration capability) of individuals in addition to their tendency to revisit positions of good memories (cognitive behavior) and tendency to keep an eye on and follow the majority of swarm members (social behavior). The balance among these three major behaviors is the key of success of the algorithm. On the other hand, there are other social and cognitive phenomena, which might be useful for improvement of the algorithm. In this paper, we particularly investigate avoidance from the bad behavior. We propose modifications about modeling the Standard PSO 2011 formulation, and we test performance of our proposals at each step via benchmark functions, and compare the results of the proposed algorithms with well-known algorithms. Our results show that incorporation of Social Avoidance behavior into SPSO11 improves the performance. It is also shown that in case the Social Avoidance behavior is applied in an adaptive manner at the very first iterations of the algorithm, there might be further improvements.