About the course:
For many industrial applications, finding a model from physical laws that is both simple and reliable for control design is a hard task. However, when a set of measurements is available, the control law can be computed from data without relying on the knowledge of the underlying physics. Specifically, in black-box model-based approaches, a model of the system is first derived from data and then a controller is computed based on such a model. In the so-called "direct" data-driven approaches, the controller is instead directly derived from experimental data. It is common belief that finding a good model of the plant is always the best way towards effective controller design, therefore model-based approaches usually prevail. In this course, an overview of data-driven methods will be provided and it will be shown that direct solutions might sometimes be preferable. Some recent results will also be discussed to highlight the potential of future research from both a theoretical and an industrial perspective.

Prerequisites:
Fundamentals of System Identification and Automatic control.

Lecture notes and code:
- Slides

- Code 

References:
- L. Ljung, System Identification: Theory for User (2nd ed.), Prentice Hall, 1999.
- M. Verhaegen and V. Verdult, Filtering and system identification: a least squares approach, Cambridge university press, 2007.
- K. Zhou, J. Doyle, and K. Glover. Robust and optimal control. Prentice hall, 1996.
- Recent scientific papers indicated in the slides.

Email to:
simone.formentin at polimi dot it