Multi-objective particle swarm-differential evolution algorithm. Issue 2 (February 2017)

• Record Type: Journal Article
• Authors: Su, Yi-xin; Chi, Rui
• Is Part Of: Neural computing & applications. Volume 28:Issue 2(2017)
• Page Start: 407
• Page End: 418
• Publication Date: 2017-02
• Subjects: Multi-objective optimization -- Particle swarm optimization -- Differential evolution -- Scale factor; Neural networks (Computer science) -- Periodicals; Neural circuitry -- Periodicals; Artificial intelligence -- Periodicals; Neural Networks (Computer) -- Periodicals; Réseaux neuronaux (Informatique) -- Périodiques; Réseaux nerveux -- Périodiques; Intelligence artificielle -- Périodiques; 006.32
• DOI: 10.1007/s00521-015-2073-y
• Languages: English
• ISSNs: 0941-0643

Abstract

Abstract A multi-objective particle swarm-differential evolution algorithm (MOPSDE) is proposed that combined a particle swarm optimization (PSO) with a differential evolution (DE). During consecutive generations, a scale factor is produced by using a proposed mechanism based on the simulated annealing method and is applied to dynamically adjust the percentage of use of PSO and DE. In addition, the mutation operation of DE is improved, to satisfy that the proposed algorithm has different mutation operation in different searching stage. As a result, the capability of the local searching is enhanced and the prematurity of the population is restrained. The effectiveness of the proposed method has been validated through comprehensive tests using benchmark test functions. The numerical results obtained by this algorithm are compared with those obtained by the improved non-dominated sorting genetic algorithm (NSGA-II) and the other algorithms mentioned in the literature. The results show the effectiveness of the proposed MOPSDE algorithm.