International Symposium on Nonlinear Theory and its Applications
Particle Swarm Optimization with Novel Concept of Complex Network
Haruna Matsushita, Yoshifumi Nishio, Toshimichi Saito,
PDF download (209.9KB)
The paper proposes a novel concept of a complex network for the Particle Swarm Optimization (PSO); Independent-minded PSO (IPSO). Particles of the standard PSO always fly toward its own past best position (pbest) and the best position among the swarm (gbest). On the other hand, in the proposed IPSO, whether each particle and the swarm affected each other is stochastically decided according to a fixed parameter called "Cooperativeness". We confirm behavior of IPSO and its effectiveness by applying it to various benchmarks.