| 粗粒度并行遗传算法性能分析 |
| |
| 引用本文： | 郑志军,郑守淇.粗粒度并行遗传算法性能分析[J].小型微型计算机系统,2006,27(6):1002-1006. |
| |
| 作者姓名： | 郑志军 郑守淇 |
| |
| 作者单位： | 1. 浙江科技学院,信息与电子工程学院,江苏,杭州,310023
2. 西安交通大学,计算机科学与技术系,陕西,西安,710049 |
| |
| 摘 要： | 依据实验来分析影响并行遗传算法性能的因素得到的结论缺乏理论上的说服力．通过对粗粒度并行遗传算法加速比公式的分析，提出了影响并行遗传算法性能的关键因素，同时否定了以迁移率作为评价并行遗传算法性能指标的合理性，并通过实难进一步验证结论的正确性．得到的结论为提高遗传算法的并行化效率提供了可靠的依据。
|
| 关 键 词： | 粗粒度 并行遗传算法 加速比 性能 分析 |
| 文章编号： | 1000-1220（2006）06-1002-05 |
| 收稿时间： | 10 10 2005 12:00AM |
| 修稿时间： | 2005-10-10 |
Performance Analysis of Coarse-Grained Parallel Genetic Algorithms |
| |
| ZHENG Zhi-jun,ZHENG Shou-qi.Performance Analysis of Coarse-Grained Parallel Genetic Algorithms[J].Mini-micro Systems,2006,27(6):1002-1006. |
| |
| Authors: | ZHENG Zhi-jun ZHENG Shou-qi |
| |
| Abstract: | So far the factors of influencing the performance of parallel genetic algorithms are always analyzed by experiments. Therefore, the conclusions generally lack convinced theoretical basis. Proposed the key factors of influencing parallel genetic algorithms by analyzing the speedup formula of coarse-grained parallel genetic algorithms. It is proved that using migration rate as its performance index is not reasonable and furthermore the correctness is proved by the experiments. The conclusion of the paper provides a reliable basis for improving the parallel efficiency of genetic algorithms. |
| |
| Keywords: | coarse-grained parallel genetic algorithms speedup pcrformance analysis |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
