Rough set based knowledge modeling for the aluminum alloy pulsed GTAW process

Abstract

The arc welding process is so complex that the classical modeling method cannot obtain the model effectively. However, the model of the arc welding process is necessary for the intelligent control of the process. Therefore, the modeling has been the interest of many researchers. Recently, more and more researchers are attempting to obtain the model of the process by means of intelligent methods, such as the neural network method, the fuzzy set method, and so on. All these methods concentrate on simulating the intelligent behavior of human beings, namely using human experience. Many applications of these methods have proved their effectiveness under certain conditions. However, their limits are obvious and further research is needed. This paper proposes a method of rough set based knowledge modeling for the aluminum alloy pulsed gas tungsten arc welding (GTAW) process. Owing to the ability of dealing with knowledge (experience) of the rough set theory, the method can obtain the knowledge model of the aluminum alloy pulsed GTAW process. The model obtained is easily understood and revised. Experiment results indicate that the method is effective. The method can be regarded as the basis of the intelligent control of the welding process.

Three-dimensional vision applications in GTAW process modeling and control

“…[24], a method of rough set-based knowledge modeling for the aluminum alloy pulsed GTAW process was proposed. Based on the passive vision robot welding system for the pulse TIG welding process, a three layer BP network model was established to predict backside width.…”

SVM-based fuzzy rules acquisition system for pulsed GTAW process

“…neural network (NN), fuzzy set (FS) and rough set (RS) (Chen et al, 1997;Chen et al, 2000a, b;Wang et al, 2005). Some researches show that the above methods can effectively obtain the model of arc welding process under certain conditions.…”

Crosscorrelation functions based nonlinear identification of diode laser welding process

“…Although it is not an easy task to model a welding process due to its nonlinearity, some progress has been made in this field. Several artificial intelligence approaches such as neural network, fuzzy set and rough set [1][2][3][4] have been proposed to obtain models for arc welding processes. In [5], a fuzzy modeling method based on Support Vector Machine (SVM) for the arc welding process is proposed.…”