Research on Measurement of Pulsed-GTAW Process by Multiple Sensors

Welding process is a non-linear, multi-variable and time-varying process accompanied by many interference factors such as dust, intense arc and electromagnetism, so it is difficult to obtain accurate information reflecting the welding process status, let alone controlling the process precisely. This paper used arc sensor, sound sensor and visual sensor simultaneously to obtain different information of the pulsed-GTAW process, and SVM was used to fuse the information to predict the quality of the welding process, experiment results showed multi-sensor information fusion could predict the process more precisely. This laid the foundation for further controlling the welding quality automatically and intelligently.