Experiments for Education in Dynamics and Control System Design
Key Features
Modular design:
Most of the PendCon experiments have a modular design. This means that many components can be used in common for different experiments.
Smart design:
The PendCon experiments have a smart and elegant design and can be easily be composed or decomposed. For this reason, many of them can be used within a lecture.
Quality of the hardware:
The mechanical and electrical are very carefully manufactured in our workshops. Third party components like encoders or motors are of highest quality bought by internationally well-known manufacturers.
Accurate simulation model:
The software associated to the experiments contains an accurate simulation model. Thus, in most cases, the controller works for the real plant if it does for the simulation model.
High-frequency dynamics
If the pendulum experiments are modeled with ideal sensors and actuators, there are situations where the simulated closed-loop systems behaves stable, whereas the controller for the real plant leads to an unstable closed-loop system. The reason is a high-frequency resonance peak in the actuator-sensor section of the plant. If this is also taken into account into the model, the mathematical model of the closed-loop system becomes also unstable in such situations. It is an impressive experience for the students, that even for apparently harmless plants unmodeled high-frequency dynamics really exists and can actually destabilize the plant. Thus, even a very good simulation model can never replace the test with the real hardware.
Uncertainty set
It is possible to change the plant by some suitable mechanical and electrical devices (they constitute the so-called uncertainty set). By this way, "uncertainty" is added to the plant which tends to destabilize the closed-loop system. This forces the students to redesign their controllers. If the uncertainty becomes too large, a redesign will not give the solution and a complete new approach is necessary.