Name
Title | ||
|---|---|---|
Sensor based weld bead geometry prediction in pulsed metal inert gas welding process through artificial neural networks |
Abstract | ||
|---|---|---|
Weld quality is primarily determined from the weld bead geometry. This work concerns the weld bead geometry prediction in pulsed metal inert gas welding (PMIGW) process. A back propagation neural network (BPNN) model, a radial basis function network (RBFN) model and regression model have been developed to predict the weld bead geometry of welded plates. Six process parameters, namely pulse voltage, back-ground voltage, pulse duty factor, pulse frequency, wire feed rate and the welding speed along with root mean square (RMS) values of two sensor signals, namely the welding current and the voltage signals, are used as input variables of the two models. The weld bead width, height and reinforcement of the welded plate are considered as the output variables. Having same process parameters does not always result in the same output quality. This is why, inclusion of sensor signals in the models, as developed in this work, results in better output prediction. |
Year | DOI | Venue |
|---|---|---|
2008 | 10.3233/KES-2008-12202 | KES Journal |
Keywords | Field | DocType |
pulsed metal inert gas,weld bead geometry prediction,output quality,process parameter,weld bead width,welding process,artificial neural network,weld bead geometry,back-ground voltage,sensor signal,better output prediction,output variable,weld quality,regression model,response surface methodology,radial basis function network | Radial basis function network,Computer science,Duty cycle,Voltage,Root mean square,Geometry,Artificial neural network,Welding,Response surface methodology,Inert gas | Journal |
Volume | Issue | ISSN |
12 | 2 | 1327-2314 |
Citations | PageRank | References |
0 | 0.34 | 4 |
Authors | ||
3 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Sukhomay Pal | 1 | 17 | 2.53 |
| Surjya K. Pal | 2 | 58 | 9.70 |
| Arun K. Samantaray | 3 | 32 | 5.37 |