A speed-up and speed-down strategy for swarm optimization

Abstract

In this paper, inspired by speed-up and speed-down (SUSD) mechanism observed by the fish swarm avoiding light, an SUSD strategy is proposed to develop new swarm intelligence based optimization algorithms to enhance the accuracy and efficiency of swarm optimization algorithms. By comparing with the global best solution, each particle adaptively speeds up and speeds down towards the best solution. Specifically, a new directed speed term is added to the original particle swarm optimization (PSO) algorithm or other PSO variations. Due to the SUSD mechanism, the algorithm shows a great improvement of the accuracy and convergence rate compared with the original PSO and other PSO variations. The numerical evaluation is provided by solving recent benchmark functions in IEEE CEC 2013.