The Development Of Online Detection System Of Resistance Spot Welding Quality Based On LabVIEW And Support Vector Machine

Resistance spot welding is widely used in industrial manufacturing fields such as automobile,aerospace and electronics,the quality of the spot has a direct influence on reliability and performance of the devices,the traditional method to detect quality of spot is the size of the nugget which can be obtained through destructive tests or shear resistance which can be obtained through tensile tests,but both methods will consume a long time and a lot of manpower and material resources,moreover,it can not realize on-line detection of the quality of the spot.A feasible method is utilize the welding parameters which indirect correlation with the quality of the spot to predict the quality of the spot.Therefore,this paper develop an online detection system of spot welding quality,which can estimate quality of the spot by using parameters that is indirect correlation with the quality of the spot,and it is of great significance to ensure the quality of the spot.This paper develops an on-line detection system of resistance spot welding quality which is based on Lab VIEW and support vector machine,it can monitors current,voltage,dynamic resistance and electrode force while welding,through the characteristic parameters which can represent the quality of the spot,the paper build a SVM model,successfully predict the quality of the spot,and store the spot information into data base,make it convenient to manage and inquire information.The main tasks of this paper are as follows:(1)Build a hardware system used for spot welding quality detection which consists of sensors,data acquisition card and signal conditioning module,realize real-time acquisition data of spot welding.(2)Use graphical programming language LabVIEW to program data acquisition,analysis,processing,storage and inquiring,realize functions like real-time display spot welding data,estimate quality of the spot and real-time store spot state.(3)Through spot welding and tensile testing,this paper extract 7 characteristic parameters which can represent the quality of the spot,build a SVM model used for spot welding quality detection,achieve a reasoning system to real-time detect quality of the spot with high accuracy by using small train specimens.