作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2012, Vol. 38 ›› Issue (19): 142-146. doi: 10.3969/j.issn.1000-3428.2012.19.037

• 人工智能及识别技术 • 上一篇    下一篇

异构系统中改进的遗传调度算法

高 原1,顾 星2,杨 群2,柯何杨2   

  1. (1. 国电南瑞科技股份有限公司研发中心,南京 210061;2. 南京航空航天大学计算机科学与技术学院,南京 210016)
  • 收稿日期:2011-11-14 出版日期:2012-10-05 发布日期:2012-09-29
  • 作者简介:高 原(1981-),男,工程师、硕士,主研方向:人工智能,调度算法,电网调度自动化技术;顾 星,硕士研究生;杨 群,副教授、博士;柯何杨,硕士研究生

Improved Genetic Scheduling Algorithm in Heterogeneous System

GAO Yuan   1, GU Xing   2, YANG Qun   2, KE He-yang   2   

  1. (1. Research and Development Center of NARI Technology Development Co. Ltd., Nanjing 210061, China; 2. College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China)
  • Received:2011-11-14 Online:2012-10-05 Published:2012-09-29

摘要: 针对以往调度算法无法准确反映任务的优先级的问题,提出一种改进的遗传调度算法。将任务的ranku属性值应用于染色体的构造方法中,即根据ranku属性值降序排列得到染色体的任务序列,使重要任务优先获得调度,从而得到优质的初始种群,为遗传算法提供合适的迭代起点。实验结果表明,该算法能有效提高初始种群的质量,获得更好的调度性能。

关键词: 任务调度, 异构计算系统, 遗传算法, 调度算法, 初始种群

Abstract: Because of the deficiencies of conventional genetic algorithm applied in task scheduling problem, this paper presents an improved scheduling algorithm based on genetic algorithm. It uses the rank up attribute values of tasks in the construction of the chromosome, in order to schedule the more important tasks preferentially, the task sequence of chromosome sorts in descending order according to rank up. Thus it can get the initial population of high quality, and the Genetic Algorithm(GA) has a good starting point for iteration. Experimental results show the algorithm can improve the quality of the population effectively, and the better scheduling performance is obtained than other conventional algorithms.

Key words: task scheduling, Heterogeneous Computing System(HCS), Genetic Algorithm(GA), scheduling algorithm, initial population

中图分类号: