A novel parallel multi-swarm algorithm based on comprehensive learning particle swarm optimization. (October 2015)

• Record Type: Journal Article
• Authors: Gülcü, Şaban; Kodaz, Halife
• Is Part Of: Engineering applications of artificial intelligence. Volume 45(2015:Sep.)
• Page Start: 33
• Page End: 45
• Publication Date: 2015-10
• Subjects: Particle swarm optimization -- Parallel algorithm -- Comprehensive learning particle swarm optimizer -- Global optimization; Engineering -- Data processing -- Periodicals; Artificial intelligence -- Periodicals; Expert systems (Computer science) -- Periodicals; Ingénierie -- Informatique -- Périodiques; Intelligence artificielle -- Périodiques; Systèmes experts (Informatique) -- Périodiques; Artificial intelligence; Engineering -- Data processing; Expert systems (Computer science); Periodicals; 620.00285
• DOI: 10.1016/j.engappai.2015.06.013
• Languages: English
• ISSNs: 0952-1976

Abstract

Abstract: This article presented a parallel metaheuristic algorithm based on the Particle Swarm Optimization (PSO) to solve global optimization problems. In recent years, many metaheuristic algorithms have been developed. The PSO is one of them is very effective to solve these problems. But PSO has some shortcomings such as premature convergence and getting stuck in local minima. To overcome these shortcomings, many variants of PSO have been proposed. The comprehensive learning particle swarm optimizer (CLPSO) is one of them. We proposed a better variation of CLPSO, called the parallel comprehensive learning particle swarm optimizer (PCLPSO) which has multiple swarms based on the master-slave paradigm and works cooperatively and concurrently. The PCLPSO algorithm was compared with nine PSO variants in the experiments. It showed a great performance over the other PSO variants in solving benchmark functions including their large scale versions. Besides, it solved extremely fast the large scale problems.