Prediction of welding parameters for pipeline welding using an intelligent system

Abstract

The determination of the welding parameters for pipeline welding is based on a skilled welder's know-how and long-term experiences rather than on theoretical and analytical techniques. In this paper, an intelligent system for the determination of welding parameters for each pass and welding position, for pipeline welding based on one database and a finite element method (FEM) model, and on two back-propagation (BP) neural network models and a corrective neural network (CNN) model was developed and validated. The preliminary test of the system has indicated that the system could determine the welding parameters for pipeline welding quickly, from which good weldments can be produced without experienced welding personnel. Experiments using the predicted welding parameters from the developed system proved the feasibility of interface standards and intelligent control technology to increase productivity, improve quality and reduce the cost of system integration.

Combining soft computing techniques and the finite element method to design and optimize complex welded products

“…In that case, models that were based on ANNs were used to predict the geometry of the weld beads in a GMAW process and, subsequently, used to build the FE models. Other authors [33] developed an intelligent system formed by a combination of FEM and ANNs to determine the welding parameters for pipeline welding. The FE models were used to calculate the pass number according to primary input welding parameters (material thickness, groove angle, material type, and wire diameter).…”

Artificial neural network modeling of weld joint strength prediction of a pulsed metal inert gas welding process using arc signals

“…are used for modeling. However, MLP, which is generally trained with the back propagation error algorithm, is popularly used in the weld modeling (Kim et al, 2006;Lee and Um, 2000;Chi and Hsu, 2001;Lightfoot et al, 2005;Ohshima et al, 1995). In this research, a code for multi-neuron, multi-hidden layer ANN model has been developed in C programming language, for mapping the pulsed metal inert gas welding (PMIGW) process parameters such as pulse voltage, back-ground voltage, pulse duration, pulse frequency, wire feed-rate, welding speed and RMS values of current and voltage signals to the ultimate tensile stress of the resulting weld joint.…”

“…In (Kang et al, 1999), an ANN model is developed to select welding parameters such as welding current, arc voltage, welding speed and weaving length for required output specifications given as weld bead shape, i.e., leg length, penetration, throat thickness and reinforcement height. An intelligent system for the automatic determination of optimal welding parameters for each pass and welding position is developed by Kim et al (2006). A finite element model, two back propagation neural network models, and a corrective neural network model are used and validated therein.…”

Application of entropy measurement technique in grey based Taguchi method for solution of correlated multiple response optimization problems: A case study in welding

“…Their developed system was capable of receiving the desired weld dimensions as input, and selecting the optimal welding parameters as outputs. Kim et al [8] applied an intelligent system for the determination of welding parameters for each pass and welding position, for pipeline welding, based on one database and a finite element method (FEM) model, and on two back-propagation (BP) neural network models and a corrective neural network model (CNN). Experiments using the predicted welding parameters from the developed system proved the feasibility of interface standards and intelligent control technology to increase productivity, improve quality and reduce the cost of system integration.…”