基于电信号多元统计的GMAW焊接质量分析方法  

Welding quality analysis of GMAW process based on multivariate statistical analysis of electric signal

作　　者：顾小燕[1] 李志勇[1] 王耀[1] 

机构地区：[1]中北大学焊接中心,太原市030051

出　　处：《焊接》2008年第12期48-51,共4页Welding & Joining

基　　金：国家自然基金资助项目(50505048);山西省青年科学基金项目(2006021027)

摘　　要：为了分析在不同干扰因素下的焊接过程信息与焊接质量的联系,应用德国汉诺威焊接质量分析仪和采集系统对GMAW焊接过程的电信息进行采集,通过对电流电压PDD曲线和短路时间CFD曲线的分析,再用聚类分析的方法对反映不同焊接过程的特征信息和五维统计矢量进行处理。实验结果表明,用多元统计的方法可以准确判断出焊接过程中不同干扰因素对焊接质量的影响。Five types of welding flaws were investigated with Hannover analyzer. Characteristic information and quintuple statistical vector reflecting welding process under the different conditions were extracted from PDD curve and electrical parameters,and analyzed with the method of clustering. The results show that multivariate statistical analysis can exactly reflect the change of the welding quality introduced by different interference factors.

关 键 词：概率密度 焊接质量 聚类分析 多元统计 

分 类 号：TG444.7[金属学及工艺—焊接]