Journal of Systems Engineering and Electronics ›› 2013, Vol. 35 ›› Issue (6): 1348-1352.doi: 10.3969/j.issn.1001-506X.2013.06.37

• 软件、算法与仿真 • 上一篇    

智能虚拟维修环境多Agent 协作机制

焦玉民 1,王强1,徐婷1,苏京1,2   

  1. 1. 解放军理工大学野战工程学院,江苏 南京 210007;
    2. 南京市住房与城乡建设委员会,江苏 南京 210016
  • 收稿日期:2012-04-16 修回日期:2012-10-19 出版日期:2013-06-15 发布日期:2013-03-29

Multi-agent cooperation mechanism for intelligent virtual maintenance environment

JIAO Yu-min1,WANG Qiang1,XU Ting1,SU Jing1,2   

  1. 1.Field Engineering Institute, PLA University of Science and Technology, Nanjing 210007, China;
    2.Nanjing Municipal Commission of Housing and Urban-Rural Development, Nanjing 210016,China
  • Received:2012-04-16 Revised:2012-10-19 Online:2013-06-15 Published:2013-03-29

摘要:

针对智能虚拟维修环境中多Agent协同作业求解问题,提出一种基于任务驱动方法的协同虚拟维修训练体系框架,并构建基于该框架下的协作感知-规划-行为(Sense-Plan-Action,SPA)关系模型,在该模型中使用参数化表达方法来量化多Agent 协同作业过程,并将任务求解过程转化为时序信息、属性状态演变、目标决策等多通道框架驱动协同作业的实现,解决了多Agent 协同作业的逻辑关系问题。最后,将该方法应用于轮式工程机械协同虚拟维修仿真过程中,实现了协同虚拟维修可视化,验证了任务驱动方法在协同虚拟训练中的可行性。

Abstract:

To solve the multi-agent cooperation problem in the intelligent virtual maintenance environment (IVTE), this paper proposes collaborative operation system based on a novel task driven method. In this method, Sense-Plan-Action (SPA) model and parametric expression are presented to quantize multi-agent cooperation process. To solve the problem of operational logical relationship, cooperation process is transformed to time-ordered information, attribute calculus, and objective decision. Finally, these methods mentioned above are applied to virtual cooperation training system of certain wheeled equipment. This system realized the visualization for collaborative operation and also illustrated the effectiveness of the proposed method.