首页 | 本学科首页   官方微博 | 高级检索  
     

一种改进的分层并行演化算法(英文)
引用本文:颜颖,曾文华. 一种改进的分层并行演化算法(英文)[J]. 心智与计算, 2007, 0(1)
作者姓名:颜颖  曾文华
作者单位:华侨大学信息科学与工程学院 福建泉州362021(颜颖),厦门大学软件学院 福建厦门361005(曾文华)
摘    要:提出了一种改进的分层并行演化算法。针对传统算法中"同构子种群"和"同步通信"所引发的问题,新算法构建了异构模型,并将各子种群充分连接。子种群一旦满足迁移条件,便可按照预设的迁移模式实施异步迁移。仿真实验结果表明,本文提出的新算法有效地解决了"征服问题"和"无效问题",避免了算法的过早收敛,提高了算法的效率。

关 键 词:分层并行演化算法  模糊交叉算子  异构模型  异步迁移

An Advanced Hierarchical Parallel Evolutionary Algorithm
YAN Ying ,ZENG Wen-hua. An Advanced Hierarchical Parallel Evolutionary Algorithm[J]. Mind and Computation, 2007, 0(1)
Authors:YAN Ying   ZENG Wen-hua
Affiliation:YAN Ying 1,ZENG Wen-hua 2
Abstract:An advanced hierarchical parallel evolutionary algorithm (AHPEA) is proposed in this paper. To solve the problems caused by "homogeneous subpopulations" and "synchronous migration" of the traditional algorithm, AHPEA builds a heterogeneous model and adequately connect subpopulations. Once a subpopulation satisfies the migration conditions, it is allowed to implement the asynchronous migration according to the predetermined migration modes. The experimental results show that, the presented algorithm efficiently solve the conquest and noneffect problems, avoid premature convergence and improve efficiency.
Keywords:hierarchical parallel evolutionary algorithm  fuzzy recombination operator  heterogeneous model  asynchronous migration
本文献已被 CNKI 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号