计算机工程与应用 ›› 2010, Vol. 46 ›› Issue (22): 217-220.DOI: 10.3778/j.issn.1002-8331.2010.22.063

• 工程与应用 • 上一篇    下一篇

车间调度中的数据挖掘技术研究

王 雯1,栾 飞1,2,傅卫平1   

  1. 1.西安理工大学 机械与精密仪器工程学院,西安 710048
    2.陕西科技大学 机电工程学院,西安 710021
  • 收稿日期:2010-01-13 修回日期:2010-02-23 出版日期:2010-08-01 发布日期:2010-08-01
  • 通讯作者: 王 雯

Study on data mining technology in workshop scheduling

WANG Wen1,LUAN Fei1,2,FU Wei-ping1   

  1. 1.School of Mechanical and Instrumental Engineering,Xi’an University of Technology,Xi’an 710048,China
    2.Mechanical & Electrical Engineering College,Shaanxi University of Science and Technology,Xi’an 710021,China
  • Received:2010-01-13 Revised:2010-02-23 Online:2010-08-01 Published:2010-08-01
  • Contact: WANG Wen

摘要: 基于Multi-Agent System(MAS)的人机合作技术适合于解决复杂调度问题。为了使人与机能够更好地合作来完成高效、准确的车间调度,引入C4.5算法,建立并实现了基于机器学习和MAS的人机合作车间调度系统仿真模型。在Java环境下,以Weka、JADE为开发平台,以Eclipse为开发工具,Access为后台数据库,完成了系统的开发。通过实例仿真和结果分析,运用机器学习算法动态调度的结果稍优于最佳的静态调度结果,证明了系统的正确性和优越性。

关键词: 数据挖掘, 机器学习, 人机合作, 车间调度, 多Agent系统

Abstract: Abstract: Man-machine cooperation technology based on Multi-Agent System(MAS) is suitable to solve complex scheduling problems.In order to better complete high efficient and accurate shop scheduling by man-machine cooperation,using the algorithm C4.5,a simulation model of man-machine cooperation system of based on machine learning and MAS technology for job shop scheduling is set up and realized.Then using Weka,JADE as the development platform,in the Java environment applying Eclipse as the development tools and Access as the background database,the development of the system has been completed.Though simulation of instances and analysis of experimental result,the results of dynamic scheduling by machine learning algorithm is slightly better than the best results of the static scheduling,which proves the validity and superiority of the system.

Key words: data mining, machine learning, man-machine cooperation, shop scheduling Multi-Agent System(MAS)

中图分类号: