首页 | 本学科首页   官方微博 | 高级检索  
     

实时分布处理系统的分层遗传调度方法
引用本文:艾锐峰,蔡志明. 实时分布处理系统的分层遗传调度方法[J]. 计算机应用研究, 2012, 29(5): 1741-1743
作者姓名:艾锐峰  蔡志明
作者单位:海军工程大学电子工程学院,武汉,430033
摘    要:提出了一种将分层分组调度算法融入实时分布处理系统的并行设计,并结合遗传算法进行任务调度的方法。首先对初始系统有向非循环图(directed acyclic graph,DAG)分层分组;然后在层间进行均衡化的乒乓流水设计,重构系统,使系统均衡化;最后针对重构的系统编制相适应的二维染色体码,运用遗传算法进行调度。实验结果表明较之单纯的分层分组方法和遗传算法,系统的时延得到明显优化,算法的收敛速度提高。

关 键 词:调度算法  并行设计  遗传算法

Layed genetic algorithm for scheduling of distributedreal-time processing system
AI Rui-feng,CAI Zhi-ming. Layed genetic algorithm for scheduling of distributedreal-time processing system[J]. Application Research of Computers, 2012, 29(5): 1741-1743
Authors:AI Rui-feng  CAI Zhi-ming
Affiliation:College of Electronic Engineering, Naval University of Engineering, Wuhan 430033, China
Abstract:This paper presentd a genetic method for scheduling the tasks of distributed real-time processing system, with joi-ning the parallel design of system and the dispatching of system jobs. First, the method partitioned the original designed directed acyclic graph of system in layers and teams. Then, it rebuilt the system with the ping-pong flow design between layers to make the whole system more proportionate. At last, it gave out a genetic sheduling algorithm for the new system with a 2D chromosome code. The experiment result shows that the delay of system reduces and the speed of this method accelerates.
Keywords:scheduling algorithm   parallel design   genetic algorithm
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《计算机应用研究》浏览原始摘要信息
点击此处可从《计算机应用研究》下载全文
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