taptap下载安装安卓学报

• 民用航空 • 上一篇    下一篇

基于协同过滤理论的民机智能故障诊断方法

薛鹏   

  1. (taptap下载安装安卓工程技术训练中心,天津300300)
  • 收稿日期:2014-03-01 修回日期:2014-04-04 出版日期:2014-08-26 发布日期:2014-10-31
  • 作者简介:薛鹏(1982—),辽宁北票人,硕士,助教,研究方向为航空发动机故障诊断.
  • 基金资助:

    中央高校基本科研业务费专项(3122014D010)

Research on fault diagnosis method of civil aircraft based on collaborative filtering theory

XUE Peng   

  1. (Engineering Techniques Training Center,CAUC,Tianjin 300300,China)
  • Received:2014-03-01 Revised:2014-04-04 Online:2014-08-26 Published:2014-10-31

摘要:

随着中国机队数量的增长,智能化故障诊断逐渐成为民航业研究的热点。协同过滤理论是在大量繁杂信息中寻找合适结论的一种有效的智能方法。将协同过滤理论应用于民机故障诊断领域,通过Pearson法和矢量余弦法计算协同过滤理论中的故障相似程度,并针对传统协同过滤算法存在的缺陷,通过引入元相似度及权重加以有效解决,从而得到更准确的结论。实例证明该方法在民机故障诊断中有较高准确率和较好的学习能力,可提高维修效率,降低运营成本。

关键词: 故障诊断, 协同过滤, 矢量余弦, 元相似度

Abstract:

Collaborative filtering theory is used in fault diagnosis field of civil aircrafts. Similarities between faults in the theory are calculated by Pearson method and vector cosine method. By analyzing the defects of collaborative filtering method,the concept of meta-similarity and weight are applied to solve these problems. Finally,practical instances in different conditions prove the learning ability and high accuracy of collaborative filtering theory and the optimized method,which can be applied in civil aircraft fault diagnosis,helping improve maintenance efficiency and reduce operating costs.

Key words: fault diagnosis, collaborative filtering, vector cosine, meta-similarity

中图分类号: 

Baidu
map