首页 | 本学科首页   官方微博 | 高级检索  
     

基于改进的遗传算法软硬件划分方法研究
引用本文:杨凯,罗胜钦. 基于改进的遗传算法软硬件划分方法研究[J]. 电子与封装, 2009, 9(1): 16-19
作者姓名:杨凯  罗胜钦
作者单位:同济大学电子与科学技术系,上海,201804;同济大学电子与科学技术系,上海,201804
摘    要:随着芯片集成度的飞速发展,集成电路的设计已经进入了片上系统(SoC,System on Chip)的时代。传统的软硬件分开设计的方法已经不再适合SoC设计的需要,而软硬件协同设计技术很好地解决了传统设计方法所不能解决的问题。软硬件划分方法是软硬件协同设计中的一个关键的问题,文章主要从基于多目标的遗传算法出发,对遗传算法主要做了两方面的改进:一方面引入小生境技术,进一步优化了算法;另一方面是引入精英保持策略,保证了算法的收敛性。并通过实验,对比不同算法之间的结果,验证了算法的收敛性。

关 键 词:SoC  软硬件划分  遗传算法  小生境技术  精英保持策略

Ameliorated Genetic Algorithm Based SoC Hardware/Software Partition Technology Study
YANG Kai,LUO Sheng-qin. Ameliorated Genetic Algorithm Based SoC Hardware/Software Partition Technology Study[J]. Electronics & Packaging, 2009, 9(1): 16-19
Authors:YANG Kai  LUO Sheng-qin
Affiliation:YANG Kai,LUO Sheng-qin (Electronic and science and technology department,Tongji University,Shanghai 201804,China)
Abstract:With the development of chip integration, the design of IC has entered into the age of SoC. And the traditional methods of hardware/sofeware partition hasn't adapt to the need of SoC design, the co-design of hardware/sofeware has settled the problem.The hardware/sofeware partition is a key question of the co-design of hardware/sofeware. The method has made two aspect improvements: one way is to import the niching methods, another way is to import elitism preserving strategy, so that it can ensure the astrin...
Keywords:SoC
本文献已被 CNKI 维普 万方数据 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