Adaptive Predictive Control of Backside Weld Width in Pulsed Gas Metal Arc Welding Using Electrical Characteristic Signals as Feedback

The control of backside weld width in pulsed gas metal arc welding (GMAW-P) is achieved with two electrical characteristic signals as feedback in the present work. The GMAW-P process is adjusted by changing the parameters of the base current period to provide the needed heat input, while the backside weld width is characterized during the amplitude-fixed peak current period. The two characteristic signals, i.e., the relative change in arc voltage and the average arc voltage during the peak current period, exhibit negative linear relationships with backside weld width and are employed as sensing signals. With these two characteristic signals as outputs, the backside weld width control system in GMAW-P is modeled and analyzed. The adaptive predictive controller is thus designed and validated in real-time control. This work provides an easily implemented way for weld geometry control in GMAW-P and thus has a practical meaning.