双因素维归约对特征库未涵盖弧焊类型的监测试验分析  

Experimental analysis of detection on arc welding types uncovered by the feature library through dimension reduction of two factors

作　　者：高理文[1,2] 薛家祥[2] 陈辉[2] 林志慧[2] 侍威[2] 

机构地区：[1]广州中医药大学医学信息工程学院,广东广州510006 [2]华南理工大学机械与汽车工程学院,广东广州510640

出　　处：《焊接技术》2013年第4期59-63,6,共5页Welding Technology

基　　金：国家自然科学基金资助项目(50875088);广东省科技计划项目(2010B010700001);番禺区科技计划项目(2010-Z-22-1);黄埔区科技计划项目(1021)

摘　　要：双因素维归约方法,可实现多种弧焊电源动特性自适应在线监测,有望破解监测技术研究落后于弧焊电源创新开发这一困局。为了检验该方法对未在系统考虑范围之内的新型弧焊电源的监测能力,本文在构建特征库时,特意不加入双脉冲焊的检测特征;然后,对双脉冲焊进行自适应监测试验。结果表明:该方法对双脉冲焊具有良好的监测能力,有力证明了它对特征库未涵盖的弧焊类型的监测性能。这使得普通的应用人员可用本方法所实现的系统对新型焊机进行监测,具有显著的现实意义。The self-adapting online detection on dynamic characteristics of multiple types of arc welding power sources can be carried out through dimension reduction of two factors.This method is expected to solve the problem that the research in monitoring technology falls behind the innovation in arc welding power supplies.As is mentioned in this paper,in order to test the effectiveness of this method in detecting new types of arc welding power sources uncovered by the system,features of double pulse welding were excluded from the feature library.Afterwards a selfadapting detection experiment was performed on double pulse welding.The results demonstrate that this method greatly facilitates the detection on double pulsed welding,which definitely proves its effectiveness in monitoring the arc welding types uncovered by the feature library.As a result,this method is of immense practical importance,for it enables ordinary users to detect new types of welding machines through the system realized by this method.

关 键 词：双脉冲 弧焊电源 维归约 监测 

分 类 号：TG444[金属学及工艺—焊接] TP274[自动化与计算机技术—检测技术与自动化装置]