基于Agent的工程保障分队任务规划研究  

Research on Mission Planning for Engineering Support Element Based on Agent

作　　者：冯柯[1] 郑伟[1] 李焕良[1] 孙伟[2] 马玉皓[1] 

机构地区：[1]解放军理工大学工程兵工程学院,江苏南京210007 [2]总参谋部陆航研究所,北京101121

出　　处：《兵工自动化》2010年第3期24-27,共4页Ordnance Industry Automation

基　　金：某型装备保障指挥仿真训练系统(4241315Q)

摘　　要：依据工程兵部分队所担负的工程保障任务,提出了一种混合Agent结构PCGFA。建立了工程保障分队Agent总体框架模型,阐述了其工程保障分队Agent的工作流程,运用自适应网络模糊推理系统(ANFIS)对工程保障分队Agent构筑指挥所掩蔽部作业进行推理、学习,分析得出工程保障分队合理的作业方式。并根据筑城分队Agent某型推土机作业记录对推理系统进行验证。实验表明,该系统具有较快的收敛速度和较好的精度,对工程保障分队任务规划方面的研究提供了一种方法和思路。According to the engineering support task, a mixed Agent architecture PCGFA was presented. Frame architecture of engineering support element agent was built and the work process of engineering support element was expounded. Adaptive neural fuzzy inference system （ANFIS） was applied to reasoning, learning and analyzing the work of building field shelter. A reasonable operating way of engineering support element was got. The inference system in this article was verified through the task records of the fortification element. The results show that the system had fast convergence rate and good precision and provided a method and thought for researching on mission planning for engineering support element.

关 键 词：工程保障分队 AGENT 自适应网络模糊推理系统(ANFIS) 任务规划 

分 类 号：TP273.2[自动化与计算机技术—检测技术与自动化装置] N945.13[自动化与计算机技术—控制科学与工程]