A self–organisation particle swarm optimisation algorithm based on L norm multi–measurements diversity feedback. (6th June 2013)

• Record Type: Journal Article
• Authors: Jiang, Shanhe; Ji, Zhicheng
• Is Part Of: International journal of computer applications technology. Volume 47:Number 2/3(2013)
• Page Start: 262
• Page End: 272
• Publication Date: 2013-06-06
• Subjects: particle swarm optimisation -- PSO -- premature stagnation -- self–organisation -- multi–measurement diversity -- L norm -- negative feedback -- population diversity -- global search -- computational efficiency -- local stagnation; Technology -- Data processing -- Periodicals; 620.00285
• Languages: English
• ISSNs: 0952-8091

Abstract

To improve the sole perception method of population diversity and premature stagnation, a self–organisation particle swarm optimisation algorithm based on L norm multi–measurements diversity feedback (SOPSO–L) is proposed, which introduces negative feedback mechanism to imitate the information interaction between the individuals. Position diversity, velocity diversity and self–cognitive diversity based on L norm are defined as perception information of the swarm. The proposed algorithm adopts multi–measurements swarm diversity as dynamic perception information to tune key parameters such as inertia weight and acceleration coefficients to make the algorithm in convergence or divergence stage. The corresponding characteristics of population diversities were studied. SOPSO–L is tested on six typical test functions and is compared to other variants of PSO presented in the literature. The results show that the proposed method not only greatly improves the global searching capability and computational efficiency, but also effectively avoids the local stagnation problem.