东北大学学报（自然科学版）

东北大学学报(自然科学版) ›› 2008, Vol. 29 ›› Issue (12): 1677-1680.DOI: -

• 论著 • 上一篇    下一篇

水压试验机故障诊断专家系统的知识获取

赵珍;胡学发;何大阔;王福利;   

  1. 东北大学信息科学与工程学院;东北大学流程工业综合自动化教育部重点实验室;
  • 收稿日期:2013-06-22 修回日期:2013-06-22 出版日期:2008-12-15 发布日期:2013-06-22
  • 通讯作者: Zhao, Z.
  • 作者简介:-
  • 基金资助:
    国家自然科学基金资助项目(60374003);;国家重点基础研究发展规划项目(2002CB312201);;流程工业综合自动化教育部重点实验室开放课题(PAL200511)

Knowledge acquisition of fault diagnosis expert system of hydraulic tube tester

Zhao, Zhen (1); Hu, Xue-Fa (1); He, Da-Kuo (1); Wang, Fu-Li (1)   

  1. (1) School of Information Science and Engineering, Northeastern University, Shenyang 110004, China; (2) Key Laboratory of Integrated Automation of Process Industry, Northeastern University, Shenyang 110004, China
  • Received:2013-06-22 Revised:2013-06-22 Online:2008-12-15 Published:2013-06-22
  • Contact: Zhao, Z.
  • About author:-
  • Supported by:
    -

RichHTML

0

PDF (PC)

513

摘要： 针对水压试验机故障诊断专家系统的知识获取问题,将减聚类算法与人机交互方法相结合,提出了一种基于减聚类的"变焦"知识获取方法.通过将知识获取问题近似为水压试验机AMESim仿真模型增压侧压力的数据分析,引入减聚类算法,并将聚类结果与专家进行交流,实现人机交互.分析了减聚类主要参数对聚类结果的影响,建立参数修改规则,根据专家评价结果不断调整减聚类参数,获得符合要求的结果.通过"变焦"方法针对某一类结果进行深入研究,从而实现知识获取.实验结果验证了方法的有效性.

关键词: 钢管, 水压试验机, 减聚类, 知识获取, 专家系统, 故障诊断

Abstract: The knowledge acquisition of fault diagnosis expert system of hydraulic tube tester is investigated, and a zooming knowledge acquisition method is put forward on the basis of subtraetive clustering algorithm which is applied to analyzing the boosting pressure acted by the tester with the results to be evaluated by experts. In this way, the parameters of subtraetive clustering algorithm can be readjusted in time according to the evaluation results, thus meeting the preset requirements. The knowledge that was acquired by " zooming " effect and investigation on a class of findings in depth is verified experimentally, and its validity is proved.

中图分类号: