• Article •    

Simulation data generation and experimental platform design for personal scheduling in workflow

ZHAO Jiong, LIU Ying-bo, REN Gen-quan, ZHANG Li   

  1. 1.Department of Computer Science and Technology, Tsinghua University, Beijing 100084, China:2.School of Software, Tsinghua University, Beijing 100084, China
  • Online:2012-08-15 Published:2012-08-25

面向工作流个人调度的仿真数据生成研究及实验平台设计

赵炯刘英博任艮全,张力   

  1. 1.清华大学 计算机科学与技术系,北京100084;2.清华大学 软件学院,北京100084

Abstract: In workflow management system, uncertainty of work items'arrival time and execution time in personal task list management was analyzed. On this basis, the simulation experiment platform for personal scheduling problem was designed and realized to compare the performance of difference scheduling algorithms. In this platform, people activity log of real workflow management system was analyzed, and time features of work items were modeled, thus the experiment data which were similar to work item in real environment were generated. Simulation experiment data were taken as input and performance difference of different algorithms on personal scheduling were obtained by optimal scheduling of multiple scheduling algorithms. The experiment was used based on real workflow data to verify the practicality of platform. The result showed that the platform could generate the task list simulated data similar to real environment, and could do comparison with personal scheduling algorithm effectively.

Key words: workflow, task scheduling, Gamma distribution, log extraction

摘要: 对工作流管理系统中个人工作列表管理问题中的工作项到达时间和执行时间的不确定性进行分析,设计并实现了针对个人调度问题的仿真实验平台,用于比较不同调度算法的优劣。该平台首先分析真实工作流管理系统中的人员活动日志,对工作项的时间特征进行建模,生成与真实环境中工作项特征一致的实验数据。然后将模拟数据作为输入,用多种调度算法进行模拟优化调度实验,得到各种算法在个人调度方面的性能差异。通过真实工作流数据的验证表明,该平台能够有效地产生与真实环境相似的工作项模拟数据,并能对个人调度算法进行有效的比较。

关键词: 工作流, 任务调度, 伽马分布, 日志提取

CLC Number: