A Novel Self-organizing Particle Swarm Optimization based on Gravitation Field Model

This paper presents a novel modified algorithm for particle swarm optimization (PSO). Initially, based on the inspiration of "universal gravitation" in nature, a kind of gravitation field model (GFM) applied to swarm intelligent optimization is designed. From the basis of GFM, a novel gravitational particle swarm optimization (GPSO) method is proposed, in which, a self-organizing field structure and the mass alterable principle are defined. The performance after a predefined number of generations of the proposed approach is validated through empirical simulations with well-known benchmarks by function optimization problem from the standard literature.