首页 | 本学科首页   官方微博 | 高级检索  
     

磨损颗粒的模糊神经网络识别研究
引用本文:任国全,张培林,李国璋,徐燕申.磨损颗粒的模糊神经网络识别研究[J].润滑与密封,2006(2):69-71,78.
作者姓名:任国全  张培林  李国璋  徐燕申
作者单位:1. 军械工程学院,河北石家庄,050003;天津大学机械工程学院,天津,300072
2. 军械工程学院,河北石家庄,050003
3. 天津大学机械工程学院,天津,300072
基金项目:中国科学院资助项目;军械工程学院校科研和教改项目
摘    要:在传统的模式识别技术中,模式分类的基本方法是利用判别函数来划分不同的类别,然而如何选择有效的判别函数以及在识别过程中如何对判别函数的参数进行修正,对于以往的模式识别技术是比较困难的。针对油液铁谱分析中磨损颗粒的识别问题,讨论了一般机械设备的磨损颗粒的特征,分析了神经网络技术和模糊数学相结合的模式,提出了基于模糊神经网络的铁谱图象分类和识别方法,分析结果表明,提出的方法对铁谱分析的智能化和快速化提供一种有效的途径。

关 键 词:铁谱分析  模糊神经网络  磨损类型  模式识别
文章编号:0254-0150(2006)2-069-3
收稿时间:2005-03-22
修稿时间:2005-03-22

Study on the Recognition of Wear Debris Based on Fuzzy Neutral Network
Ren Guoquan,Zhang Peilin,Li Guozhang,Xu Yanshen.Study on the Recognition of Wear Debris Based on Fuzzy Neutral Network[J].Lubrication Engineering,2006(2):69-71,78.
Authors:Ren Guoquan  Zhang Peilin  Li Guozhang  Xu Yanshen
Affiliation:1. Ordnance Engineering College, Shijiazhuang Hebei 050003, China; 2. College of Mechanical Engineering,Tianjin University,Tianjin 300072, China
Abstract:The basic analysis method in the traditional model recognition technique is to use differentiating function, however,it is difficulty to choose the differentiating function and modify its parameters in the recognition procedure. The wear debris character of the ordinary machine was discussed according to the problem of wear debris kinds in ferrography, and the combing model of neutral network and fuzzy math was analyzed. A classifying and recognition method of ferrography image based on fuzzy neutral network technique was put forward. The experiment result proves the efficiency of the method, and it is an intelligent and fast method for the ferrograph analysis.
Keywords:ferrograph analysis  fuzzy neutral network  wear model  model recognition
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号