Projects

The motivation is to experience a more complete procedure than what we can have within tightly focused weekly homework problems. It is an excellent opportunity to apply the knowledge and skills acquired during the course to a more complex problem.

Organization

  • Team work allowed, in fact even encouraged, teams of 3-5. One suggestion for the organization of the team work is that every team member can design a different kind of a controller than the other members of the team. For instance, one member can design a reference tracking LQR, another one can do an MPC and yet another can try a robust controller using Hinf optimization. The benefit of working in a team is that you can share the unavoidable pain (or fun) of learning to operate the particular laboratory experiment and possibly making sense of its model. Later you can also share the load of writing the report.

  • Each team or individual has to submit a report organized as follows:

    • A very brief description of the laboratory model and the mathematical model you used. No need to replicate the full description provided in the model repositories, but the description should be sufficient for the report to be understandable without the need to consult the model repository. In particular, included should be the explicit enumeration of all the input, state, and output variables, their ranges, units, some characterization of uncertainty in the model and the measurements. The requirements on the control system shoud be clearly stated.
    • Documentation of the computational procedure(s) leading to the design of controllers.
    • Discussion of the performance of the designed controllers with the aid of figures showing and comparing the simulated and measured responses.

  • It is highly recommendable, that each project has their own repository within the gitlab.fel.cvut.cz. The final report (6 pages at maximum) should then refer to the content (scripts, models, data) of the repository to support reproducibility.

Time

While the project can be expected to demand higher time investment then a weekly homework, it should not be another diploma thesis. Admittedly, some of the assignments rather are open-ended, which means that it is perfectly appropriate to upper bound the total time invested into the project – just do what you can within the time. But then include the timing information in the final report.

Assessment

The final report will be evaluated as either sufficient or insufficient, it will be either accepted or not. We will not assess contributions of individual students to the report, we will consider the report as a whole. There will be two attempts to submit the report, so that even if the report fails to be accepted when submitted for the first time, some feedback will be given by the instructor and there will be one more chance to submit the report.

Proposed projects

The proposed project consists of laboratory experimental systems in our KN:E-26 laboratory. The essence of the projects is to demonstrate one or preferrably several control design methods using these systems.

Quanser Active Suspension

Gitlab repository: https://gitlab.fel.cvut.cz/aa4cc/edu-ctrl-lab-models/quanser-active-suspension

Quanser Active Mass Dampers on Shake Table

Gitlab repository: https://gitlab.fel.cvut.cz/aa4cc/edu-ctrl-lab-models/quanser-amd

Quanser 3DOF Helicopter

Gitlab repository: https://gitlab.fel.cvut.cz/aa4cc/edu-ctrl-lab-models/quanser-3dof-helicopter

Back to top