计算机科学 ›› 2011, Vol. 38 ›› Issue (7): 190-193.
• 人工智能 • 上一篇 下一篇
程适，史玉回
出版日期:
发布日期:
基金资助:
CHENG Shi,SHI Yu-hui
Online:
Published:
摘要： 提出了一种新的基于L:范式的粒子群算法群体多样性定义，这种观测方式可以准确地描述算法运行过程中的信息。首先，通过对比新的观测方式和已有方式，解释了新的观测方式的特点;然后通过实验观测了位置、速度和认知三种群体多样性在算法执行过程中的变化，给出了群体多样性的变化特征。最后讨论了粒子群算法在不同解空间维数、不同粒子群拓扑结构和不同粒子数目时的群体多样性的变化情况。
关键词: 演化计算，粒子群优化，群体多样性，位置多样性，速度多样性，认知多样性，范式
Abstract: A novel PSO population diversity based on norm was defined, which provides useful information of PSO search process. Population diversity based on and norms were analyzed as well as element wised and dimension-wised PSO diversity. Population diversities of PSO with different number of dimensions,different topology structure,and different population sizes were discussed and tested on benchmark functions.
Key words: Evolutionary computation, Particle swarm optimization, Population diversity, Position diversity, Velocity diversity, Cognitive diversity, Norm
程适，史玉回. 基于L1范式的粒子群算法群体多样性研究[J]. 计算机科学, 2011, 38(7): 190-193. https://doi.org/
CHENG Shi,SHI Yu-hui. Measurement of PSO Diversity Based on L1 Norm[J]. Computer Science, 2011, 38(7): 190-193. https://doi.org/
0 / / 推荐
导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks
链接本文: https://www.jsjkx.com/CN/
https://www.jsjkx.com/CN/Y2011/V38/I7/190
Cited
首页