Analysis of vector evaluated particle swarm optimization guided by non-dominated solutions: Inertia weight, cognitive, and social constants
Recently, an improved Vector Evaluated Particle Swarm Optimization (VE PSO) algorithm has been introduced by redefining the swarm’s leader as non-dominated solutions. The improved VEPSO algorithm is named as VEPSO-non-dominated-solution (VEPSOnds). Since a parameter tuning of a heuristic algorithm i...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| 格式: | Article |
| 出版: |
2015
|
| 主题: | |
| 在线阅读: | http://eprints.utm.my/57814/ http://eprints.utm.my/57814/ |
| 标签: |
添加标签
没有标签, 成为第一个标记此记录!
|
| 总结: | Recently, an improved Vector Evaluated Particle Swarm Optimization (VE PSO) algorithm has been introduced by redefining the swarm’s leader as non-dominated solutions. The improved VEPSO algorithm is named as VEPSO-non-dominated-solution (VEPSOnds). Since a parameter tuning of a heuristic algorithm is normally difficult, in this paper, three important parameters of the improved VEPSO, which are inertia weight, cognitive constant, and social constant, are analyzed. The results suggest that the inertia weight should gradually degrade from 1.0 to 0.4, and both cognitive and social constants are random value in between 1.5 and 2.5. Analysis of vector evaluated particle swarm optimization guided by non-dominated solutions: Inertia weight, cognitive, and social constants. |
|---|