Discrete multi-phase particle swarm optimization

被引：16

作者：

Al-kazemi, B ^{[1
]}

Mohan, CK ^{[1
]}

机构：

[1] Syracuse Univ, Dept EECS, Syracuse, NY 13244 USA

来源：

INFORMATION PROCESSING WITH EVOLUTIONARY ALGORITHMS: FROM INDUSTRIAL APPLICATIONS TO ACADEMIC SPECULATIONS | 2005年

关键词：

D O I：

10.1007/1-84628-117-2_20

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

This chapter proposes the Discrete Multi-Phase Particle Swarm Optimization (DiMuPSO) algorithm, extending the PSO approach to problems coded with discrete binary representations. The main features of DiMuPSO are in utilizing multiple groups of particles with different goals that are allowed to change with time, alternately moving toward or away from the best solutions found recently. DiMuPSO also enforces steady improvement in solution quality. accepting only moves that improve fitness. Experimental simulations show that DiMuPSO outperforms a genetic algorithm and a previous discrete version of PSO on several benchmark problems.

引用

页码：305 / 327

页数：23