首页 | 本学科首页   官方微博 | 高级检索  
     检索      

别克轿车副车架总成焊接漏焊的智能检测方法
引用本文:林涛,赵学,王兵.别克轿车副车架总成焊接漏焊的智能检测方法[J].上海交通大学学报,2001,35(3):481-484.
作者姓名:林涛  赵学  王兵
作者单位:上海交通大学材料科学与工程学院,
基金项目:海市汽车工业科技发展基金(2328A)资助项目;并受"长江学者奖励计划"资助
摘    要:提出了一套漏焊检测系统的硬件设计和软件实现方案。硬件设计关键主要是通过大量数据采样实验对采样频率和数字滤波方案的确定。手工焊添丝量的波动范围很大,软件不能仅通过添丝量来检测漏焊,设计集中在通过BP网络对一焊接过程内的数次断弧进行特征提取以建立知识库,然后根据知识库的中的知识来划分不同焊缝。软件在运行中可根据运行记录库自学习优化知识库中的知识。

关 键 词:漏焊检测  特征提取  副车架  别克轿车  智能检测系统  知识库
文章编号:1006-2467(2001)03-0481-04
修稿时间:2000年6月6日

Intelligent Detecting System for Lack of Weld of BUICK Car's Subframe
LIN Tao,ZHAO Xue,WANG Bing.Intelligent Detecting System for Lack of Weld of BUICK Car's Subframe[J].Journal of Shanghai Jiaotong University,2001,35(3):481-484.
Authors:LIN Tao  ZHAO Xue  WANG Bing
Abstract:The hardware design and the project of software in an intelligent detecting system for lack of weld were proposed. The key hardwere design is to choose the sampling frequency and method of numerical filter, which is on the basis of a large number of sampling experiment data. Because the filler metal quantity of manual welding violates acutely, the software can not judge the lacking of weld by filler metal quantity only. The design focuses on the feature extracting in a welding course by the BP neural net to build knowledge library, then distinguishing every weld according to the knowledge in knowledge library. The software may self study while running to optimize the knowledge in knowledge library.
Keywords:detecting the lack of weld  feature extracting  knowledge  subframe
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