计算机集成制造系统 ›› 2021, Vol. 27 ›› Issue (8): 2282-2294.DOI: 10.13196/j.cims.2021.08.011

• 当期目次 • 上一篇    下一篇

云边协同计算架构下大规模工厂接入的任务调度

满君丰,赵龙乾,彭成+,李倩倩   

  1. 湖南工业大学计算机学院
  • 出版日期:2021-08-31 发布日期:2021-08-31
  • 基金资助:
    国家重点研发计划重点专项资助项目(2018YFB1003401);国家自然科学基金资助项目(61871432,61702177);湖南省自然科学基金资助项目(2018JJ4063,2019JJ60008,2017JJ3065,2020JJ6086,2020JJ4275);湖南省教育厅科研资助项目(16A059,17A052);湖南省研究生创新基金资助项目(CX2018B740,CX20190847)。

Task scheduling method for large-scale factory access in cloud and edge collaborative computing architecture

  • Online:2021-08-31 Published:2021-08-31
  • Supported by:
    Project supported by the National Key Research & Development Program,China(No.2018YFB1003401),the National Natural Science Foundation,China(No.61871432,61702177),the Natural Science Foundation of Hunan Province,China(No.2018JJ4063,2019JJ60008,2017JJ3065,2020JJ6086,2020JJ4275),the Hunan Provincial Department of Education Scientific Research Program,China(No.16A059,17A052),and the Graduate Innovation Foundation of Hunan Province,China(No.CX2018B740,CX20190847).

摘要: 针对大型制造业企业生产车间业务流程复杂,传统的任务调度方法无法适配云边协同的业务模式,导致负载不均衡的问题,提出云边协同任务调度算法。首先将云端服务器和边缘端服务器上的有向无环图(DAG)合并,然后采用基于关键路径的分割策略划分任务,以有效提高分配的准确性,最后通过合理分配处理器实现负载均衡。实验结果表明,所提方法通过提高计算资源利用率来缩短处理时间,并采用异构分布式系统使得处理大规模任务集时仍具有较高的计算性能。

关键词: 云制造, 云边协同, 任务调度, 有向无环图合并, 有向无环图分割, 处理器分配

Abstract: Large-scale manufacturing enterprises have complex business processes in their production workshops,and traditional task scheduling methods cannot adapt to the cloud and edge collaborative business model,which leads to the problem of unbalanced load.Aiming at this problem,a Cloud and Edge Collaborative Task Scheduling (CECTS) algorithm was proposed.The method merged Directed Acyclic Graph(DAG) on cloud server and edge server,then the tasks was divided by critical path-based segmentation strategy,which could improve the accuracy of allocation effectively.Through processor allocation,the load balancing was achieved.Experimental results showed that the method reduced the processing time by improving the utilization of computing resources,and used a heterogeneous distributed system to achieve high computing performance when processing large-scale task sets.

Key words: cloud manufacturing, cloud and edge collaboration, task scheduling, directed acyclic graph merger, directed acyclic graph segmentation, processor allocation

中图分类号: