Ruven Weiss

Friday, November 20, 2026, 12:00 - 14:00

SR 02-016/18, Geb. 101

This thesis deals with edge banding, a process in furniture production where a strip-like material is bonded to the narrow face of a panel-shaped workpiece (typically particleboard) after it has been cut to size.
Therefore, the material is heated to a specific temperature (bonding temperature) to ensure optimal adhesion.
The heating is achieved by moving it past a hot air nozzle.
Dynamic changes in transportation speed, which are inherent to the process render temperature control particularly challenging.

The thesis presents a predictive control approach allowing to adjust the air flow through the nozzle in advance of changes in transportation speed.
It is shown both in simulation and real-world experiments that the developed method is largely superior over the status quo method currently used in the industry.

Therefore, the process is mathematically modeled for the first time, and a simplified model is derived that is computationally efficient for real-time use while still providing sufficient accuracy in predicting the bonding temperature.
In this context, it is investigated which physical phenomena can be neglected and which must be modeled.
An early-lumping approach is then applied, in which the partial differential equation is transformed into a system of ordinary differential equations by discretizing it with finite differences on a tailored grid.
System Identification is used to calibrate to the real-world process.

Based on this model, a predictive control strategy is developed that exploits the specific properties and structure of the system.
The system structure is analyzed in detail, and its differences from classical systems are highlighted.

To ensure real-time capability on standard industrial hardware, an efficient optimization algorithm is developed that takes advantage of the system's specific structure.
Its runtime performance is compared to an ad hoc implementation using a state-of-the-art, but off-the-shelf solver.

Finally, real-world experiments are conducted, demonstrating the effectiveness of the proposed method for temperature control.