Cyber-Physical Systems (CPS) are complex engineering systems integrating physical processes with computation, communication and control. CPS are typically composed by a number of spatially distributed system components and/or subsystems exchanging information through communication networks. The inherent distributed and networked structure of CPS poses new control design challenges. As a matter of fact, the designed feedback control loops has to guarantee robustness against (i) uncertainty in the mathematical model of the physical process, (ii) adversarial network attacks and (iii) networked-induced delays and perturbations. Furthermore, decentralized control design techniques has to exploited in order to fulfill global performance requirements of CPSs by exploiting sensor information locally available in each subsystem. The aim of the proposed course is to provide the students with tools for modeling and control of CPSs.

At the end of the course, the student will know the main paradigms of CPSs and will be able to design and control a basic CPS. Specifically, she/he will learn how to: - mathematically model a CPS; - describe and predict the behavior of a CPS; - control the behavior of a CPS; - implement the theoretical notions in real-world CPS applications.

Linear Algebra, Linear system theory, Automatic Control, basic notions of Convex Optimization.

- Introduction to CPSs: general definition and motivations. Real-world examples of CPSs. - Modeling paradigm for CPSs: introduction to distributed networked dynamical systems. - Modeling of the physical process part of CPSs by means of Set-membership system identification techniques. Characterization of model uncertainty. - Analysis of critical aspects of distributed networked dynamical systems: adversarial network attacks, networked-induced delays and perturbations. - Review of observer design for linear systems. - Review of observer-based output feedback regulation of linear systems. - Sparse signal recovery. - Distributed/decentralized control (DC): observer-based decentralized control structure. - Networked control systems: stability in presence of networked-induced delays and perturbations. - Secure estimation and control of CPS in presence of sensor and actuator networks signals attacks: motivation and problem formulation. - Output regulation approach to secure estimation and control of CPS based on sparse signal recovery.

The course consists of a theoretical part (mathematical modeling, optimization, theory of distributed systems) alternated with practical examples and implementation. For each theoretical concept, related real-world case studies are proposed (e.g., smart grids, autonomous vehicles, distributed manufacturing systems).

Since the considered topics are quite recent, main learning resources are: - Slides and lectures notes provided by the teachers - Selected scientific papers Supplementary material: - some parts of the course refer to the book Cyber-Physical Systems: From Theory to Practice, D. B. Rawat, J. J.P.C. Rodrigues, I. Stojmenovic, 1st ed. 2016, CRC Press - for a gentle introduction to CPSs: Industry 4.0: The Industrial Internet of Things, A. Gilchrist, 2016, Apress

Modalità di esame: Prova scritta tramite PC con l'utilizzo della piattaforma di ateneo; Elaborato progettuale in gruppo;

The exam consists of two parts. In the first part students are required to answer questions about the entire content of the course. The maximum score for this part is 20 points. in the second part a group project activity is assigned in order to check the ability of the students to apply the tools and results presented in the lectures to real-world case studies. The maximum score for this part is 10 points.

Modalità di esame: Prova scritta (in aula); Prova scritta tramite PC con l'utilizzo della piattaforma di ateneo; Elaborato progettuale in gruppo;

The exam consists of two parts. In the first part students are required to answer questions about the entire content of the course. The maximum score for this part is 20 points. in the second part a group project activity is assigned in order to check the ability of the students to apply the tools and results presented in the lectures to real-world case studies. The maximum score for this part is 10 points.