A fresh Particle Swarm Optimizations: A position paper

This paper contributes a novel Particle Swarm Optimization (PSO) method. The particle is updated not only by the best position in history (p best ) and the best position among all the particles in the swarm (g best ), but also using the position that is nearest neighbor of p best . Additionally, we introduce a modified PSO algorithm based on the fuzzy clustering of particles to communication with the nearest neighbor for reducing the premature convergence and in sequel enhance the capability of global exploration. We validate our methods by an extensive experimental study on four benchmark test functions and compare the result with basic PSO.