An improved self-adapting Glowworm Swarm Optimization algorithm

Glowworm Swarm Optimization (GSO) is a novel heuristic algorithm based on swarm intelligence by simulating behavior of glowworms. Its advances include relatively low computational complexity and fast convergence. However, basic glowworm swarm optimization is vulnerable to local optimum. This paper presents an improved self-adapting glowworm swarm algorithm to solving such problem. By adapting radial sensor range with iteration time, new algorithm´s local searching ability is enhanced, therefore decreases probability of being trapped in local optimum. Experiments based on five standard testing multi-peak functions were performed, and the superiority of new GSO was demonstrated.