| 引用本文: | 陈保娣,曾建潮.改进的吸引扩散微粒群算法[J].控制理论与应用,2010,27(4):451~456.[点击复制] |
| CHEN Bao-di,ZENG Jian-chao.Modified attractive and repulsive particle swarm optimization[J].Control Theory and Technology,2010,27(4):451~456.[点击复制] |
|
| 改进的吸引扩散微粒群算法 |
| Modified attractive and repulsive particle swarm optimization |
| 摘要点击 1187 全文点击 1088 投稿时间：2008-12-10 修订日期：2009-04-26 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 |
| 2010,27(4):451-456 |
| 中文关键词 微粒群算法 种群多样性 微粒最好飞行方向 收敛 |
| 英文关键词 particle swarm optimization swarm diversity particle’s best flight direction convergence |
| 基金项目 教育部重点科研资助项目(204018); 山西省自然科学基金资助项目(2008011027-2). |
|
| 中文摘要 |
| 为了避免微粒群算法存在的过早收敛问题, 在ARPSO的基础之上, 提出了一个简单的种群多样性度量函数和微粒最好飞行方向的概念, 引入了变异策略, 从而实现了一种改进的吸引扩散微粒群算法MARPSO, 并从理论上分析了MARPSO的局部收敛性和全局收敛性. 对四个经典函数进行了仿真测试, 测试结果表明: 与基本微粒群算法BPSO和ARSPO相比, 该算法能够有效的提高种群多样性, 并且具有较高的收敛速度. |
| 英文摘要 |
| To avoid the premature convergence, based on the attractive and repulsive particle swarm optimizer(ARPSO), we propose a novel measure function for the population diversity, and a new concept of the particle’s best flight direction. A modified ARPSO(MARPSO) is proposed by introducing a mutation strategy. Moreover, theoretical analysis has been made to prove that the algorithm can guarantee local convergence and global convergence. By comparing the simulation results of four classic testing functions with basic PSO(BPSO), ARPSO and MARPSO, this algorithm shows an effective increase in the diversity of population, and the improvement of convergence speed. |
|
|
|
|
|