• 论文 •    

基于混合结构树的故障诊断技术研究

孙  宇,彭  强,张晓阳,陆宝春   

  1. 南京理工大学 机械工程学院, 江苏  南京  210094
  • 出版日期:2005-07-15 发布日期:2005-07-25

Fault diagnosis technology based on hybrid structure tree

SUN Yu, PENG Qiang, ZHANG Xiao-yang, LU Bao-chun   

  1. Sch. of Mechanical Eng., Nanjing Univ. of Sci. & Tech., Nanjing  210094, China
  • Online:2005-07-15 Published:2005-07-25

摘要: 根据产品结构树和故障树提出了混合结构树的概念和相应的智能故障诊断方法。该方法将诊断问题中一些基本的、具有共性的内容构成系统内核,并与具体诊断对象相区分,使其适用不同的应用对象;同时针对智能诊断中的知识管理,建立了一套编码系统,使知识管理和应用能利用现有编码管理的先进方法,实现了诊断知识的有效管理和利用,最终完成了相应诊断推理算法设计。最后,以高速列车某型号气动塞拉门为实例进行了研究和应用系统开发,现已成功应用于企业客户服务中。

关键词: 混合结构树, 故障诊断, 专家系统, 列车自动门

Abstract: Based on structure tree and fault tree of complex product, the concept of hybrid structure tree and the corresponding intelligent fault-diagnosis method were proposed. Some basic contents, which shared similar characteristics, were constructed as diagnosis kernel, and these contents were differentiated from concrete diagnosis objectives so as to make it suitable for different objects. Aimed at knowledge management in intelligent diagnosis, a set of coding rules was proposed to implement effective management and utilization of diagnosis knowledge by making use of existing advanced coding management methods. Finally, the corresponding inference algorithms of fault diagnosis were developed. All above-mentioned methods have been applied in research and development of the fault-diagnosis system for automatic door of high-speed train.

Key words: hybrid structure tree, fault diagnosis, expert system, train automatic door

中图分类号: