An adaptive particle swarm optimization within the conceptual framework of computational thinking

Abstract

The individual learning and team working is the quintessence of particle swarm optimization (PSO). Within the conceptual framework of computational thinking, the every particle is seen as a computing entity and the whole bird community is a generalized distributed, parallel, reconfigurable and heterogeneous computing system. Meanwhile, the small world network provides a favorable tool for the topology structure reconfiguration among birds. So a learning framework of distributed reconfigurable PSO with small world network (DRPSOSW) is proposed, which is supposed to give a systemative approach to improve algorithms. Finally, a series of benchmark functions are tested and contrasted with the former representative algorithms to validate the feasibility and creditability of DRPSOSW.