Application of neural networks for quality evaluation of resistance spot welds

被引：18

作者：

Dilthey, U ^{[1
]}

Dickersbach, J ^{[1
]}

机构：

[1] Univ Aachen, ISF Welding Inst, D-52062 Aachen, Germany

来源：

ISIJ INTERNATIONAL | 1999年 / 39卷 / 10期

关键词：

resistance spotwelding; quality evaluation; neural networks; steel; electrode wear; parameter variations;

D O I：

10.2355/isijinternational.39.1061

中图分类号：

TF [冶金工业];

学科分类号：

0806 ;

摘要：

A neural network based system was developed to perform non-destructive quality evaluation of spot welds. The spot welds are classified with a high reliability of 100%, taking variations of the welding parameters and electrode wear into account. The classification was carried out by the means of preprocessed data from the course of welding current and voltage.

引用

页码：1061 / 1066

页数：6