To ensure the competitiveness of manufacturing companies need to optimize their manufacturing processes. Slight variations of process parameters and machine settings only have marginally effects on the product quality. Therefore, the largest possible editing window is required. Such parameters are, for example, the movement of the laser beam across the component for the laser keyhole welding. That’s why it is necessary to keep the formation of welding seams within specified limits. Therefore, the quality of laser welding processes is ensured, by using post-process methods, like ultrasound inspection, or special in-process methods. These in-process systems only achieve a simple evaluation which shows whether the weld seam is acceptable or not. This functionality is shown graphically in Figure 1. Furthermore, in-process systems use no feedback for changing the control variables such as speed of the laser or adjustment of laser power.
In this paper the ERDF-funded research group INTEGRATiF presents current results of the research field of Online Monitoring and Model predictive controlling in laser welding processes to increase the product quality. To record the characteristics of the welding process, tested online methods are used during the process. Based on the measurement data, a state space model is ascertained, which includes all the control variables of the system. Depending on simulation tools the model predictive controller (MPC) is designed for the model and integrated into an NI-Real-Time-System.