计算机集成制造系统 ›› 2016, Vol. 22 ›› Issue (第2期): 372-380.DOI: 10.13196/j.cims.2016.02.010

• 产品创新开发技术 • 上一篇    下一篇

时间约束云工作流调度的粒子群搜索方法

曹斌,王小统,熊丽荣,范菁+   

  1. 浙江工业大学计算机科学与技术学院
  • 出版日期:2016-02-29 发布日期:2016-02-29
  • 基金资助:
    国家自然科学基金资助项目(61173097,61202202);浙江省重大科技专项重大工业资助项目(2013C01112);杭州市重大科技创新专项资助项目(20132011A16)。

Searching method for particle swarm optimization of cloud workflow scheduling with time constraint

  • Online:2016-02-29 Published:2016-02-29
  • Supported by:
    Project supported by the National Natural Science Foundation,China(No.61173097,61202202),the Major Science and Technology Project of Zhejiang Province,China(No.2013C01112),and the Major Science and Technology Innovation Project of Hangzhou City,China(No.20132011A16).

摘要: 为了快速找到较优的调度方案,针对时间约束工作流调度问题,即能在满足用户的截止时间约束的条件下最小化调度费用,提出基于粒子群算法的最优调度方案搜索方法。利用关键路径进行粒子初始化和搜索阶段的筛选处理,不但能够显著提高搜索结果的精度,而且减少了搜索的计算时间。将改进算法和传统粒子群优化算法进行了实验评估对比,实验数据证明,使用该方法使粒子搜索的时间少于传统粒子群算法,并且结果也优于传统方法。

关键词: 工作流调度, 粒子群算法, 关键路径, 云计算

Abstract: To get the optimal scheduling scheme quickly,aiming at the problem of scheduling time constrained cloud workflow which was minimized the total computing cost under the user's temporal constraint,a searching method of optimal scheduling solution based on Particle Swarm Optimization (PSO) algorithm was proposed.Owing to the problems such as random particles of traditional PSO algorithm and low search speed during the iteration,the critical path was used to select particle initialization and search space during the iterations,which was not only improved the searching accuracy,but also reduce the computing time.Compared with the traditional PSO,the experiment results proved that the proposed approach could achieve better performance with less computation time.

Key words: workflow scheduling, particle swarm optimization, critical paths, cloud computing

中图分类号: