Artificial neural networks applied to arc welding process modeling and control

Andersen, Kristinn; Cook, Gerald; Karsai, Gábor; Ramaswamy, K.

doi:10.1109/28.60056

Cited by 157 publications

(45 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The performance of the deep regression model is compared in contrast to a number of selected studies from the literature namely, Anderson et al, [7] Chandrasekhar and Vasudevan, [21] and Karaoglu and Secgin. [4] These articles form a reliable representatives of the trends of research in this field over the past two and a half decades.…”

Section: Case Studymentioning

confidence: 99%

“…[4] These articles form a reliable representatives of the trends of research in this field over the past two and a half decades. More specifically, Anderson et al [7] is the seminal paper that introduces the application of multi-layer neural networks in estimation of the direct weld parameters. Furthermore, Karaoglu and Secgin [4] apply a curvilinear approach to estimation of DWP.…”

Section: Case Studymentioning

confidence: 99%

“…Additionally, the welding data of these studies are publicly available through their respective articles. They pertain to three main welding techniques, namely arc welding, [7] submerged arc welding, [4] and A-TIG weldings on 304LN and 316LN stainless steel. [21] Throughout this Section, the following abbreviations are used to refer to each of these datasets:…”

Section: Case Studymentioning

confidence: 99%

“…These models are successfully deployed in investigation of the modeling, the process control, and the estimation of the quality of the weld beads. [22] Anderson et al [7] use ANN to model the arc welding process. Their model achieves satisfactory results on its prediction accuracy as compared to the conventional control systems.…”

Section: Introductionmentioning

confidence: 99%

“…This explains the existence of a wide range of welding techniques from Gas Metal Arc Welding (GMAW) [1] and Submerged Arc Welding (SAW) [2][3][4] to shielded metal-arc welding (SMAW), [5] Tungsten Insert Gas Welding (TIG), [6] and Gas Tungsten Arc Welding (GTAW). [7] refers to the study of the effect of the changes of IWP on the final weld quality. Whereas, the estimation problem attempts to predict the weld parameters, given the IWP.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Multi-block deep regression with reinforced gradients: Modelling the estimation of direct weld parameters in automated welding

Keshmiri

2016

AIR

View full text Add to dashboard Cite

This paper presents a deep regression model to estimation of the weld bead parameters in welding tasks. It is an aggregate of deep regression blocks where number of these blocks is proportional to the cardinality of the weld parameters. These blocks are trained simultaneously and share an identical structure with four-hidden-layer Sigmoid activation functions and a linear transformation at their outputs. Moreover, they incorporate a new meta-parameter, shared by all the hidden layers of a given block, to maintain the quality of the gradients of their respective weight matrices. This allows the model to further reduce the deviation of its estimates from the expected values of the weld parameters to significantly minimize its estimation error. The evaluation of the performance of this approach in contrast to state-of-the-art techniques in the literature shows a significant improvement in estimating these values for different welding processes. Furthermore, the proposed deep regression network is capable of retaining its performance when presented with combined data of different welding techniques. This is a nontrivial result in attaining an scalable model whose quality of estimation is independent of adopted welding techniques.

show abstract

Section: Case Studymentioning

confidence: 99%