Control Engineering

The course is organized in two modules: the first deals with the design of the so-called intelligent control systems, whereas the second will cover hybrid systems. The first module presents the basic methods for the design of intelligent control systems based on data-driven techniques such as deep learning and reinforcement learning. The course will present both the role of machine learning in control systems, discussing its methodologies, potential and applications, and the design of control systems based on artificial intelligence techniques. The course will also focus on how recent developments in deep learning can be exploited by standard control systems, discussing the basics of advanced data analysis using deep neural networks in the context of classical automation problems such as quality control, predictive maintenance and disturbance rejection. In addition to the theoretical aspects, applications from the domains of industrial automation, robotics and cyber security will be presented. At the end of the module, the student will possess the basic knowledge that will allow her/him to analyse and design intelligent systems capable of controlling complex, nonlinear processes. The second module introduces the student to the area of hybrid systems, that is dynamical systems characterized by the interaction of different types of dynamics, both continuous and discrete. The systematic study of hybrid systems is required by recent technological innovations, which led to the pervasive diffusion of increasingly complex digital systems for the control and supervision of physical systems. The study of hybrid systems is generally more challenging than that of purely discrete or purely continuous systems, because of the interaction between dynamics of different nature. Models for hybrid systems will be introduced in this course and general methods to investigate their properties will be described. Control of hybrid systems will also be addressed by focusing on some case studies from different application contexts. Students attending the course should, at the end of the second module, be able to appreciate the diversity of phenomena that arise in hybrid systems, and understand how concepts that are classical in the theory of discrete event systems, modeled by automata, can coexist with concepts that are classical in the theory of continuous systems, modeled by differential or difference equations, in a unifying framework.

- System Theory - Automatic Control - Fundamentals of statistics and probability - Fundamentals of optimal control

Sutton RS, Barto AG. Reinforcement learning: An introduction. MIT press; 2018. John Lygeros, Lecture Notes on Hybrid Systems, 2004

Classroom attendance is not mandatory.

The final grade will be based on a project presentation about one of the two modules and a written test about the other. The project has to be an independent achievement, where the goals and contents are defined by each student in agreement with professors.

Frontal teaching.

The course is organized in two modules: the first deals with the design of the so-called intelligent control systems, whereas the second will cover hybrid systems. The first module presents the basic methods for the design of intelligent control systems based on data-driven techniques such as deep learning and reinforcement learning. The course will present both the role of machine learning in control systems, discussing its methodologies, potential and applications, and the design of control systems based on artificial intelligence techniques. The course will also focus on how recent developments in deep learning can be exploited by standard control systems, discussing the basics of advanced data analysis using deep neural networks in the context of classical automation problems such as quality control, predictive maintenance and disturbance rejection. In particular, great focus will be given to application examples from the domains of industrial automation, robotics and cyber security where intelligent control solutions may contribute positively to the robustness and the economic and environmental sustainability of the controlled system. At the end of the module, the student will possess the basic knowledge that will allow her/him to analyse and design intelligent systems capable of controlling complex, nonlinear processes. The second module introduces the student to the area of hybrid systems, that is dynamical systems characterized by the interaction of different types of dynamics, both continuous and discrete. The systematic study of hybrid systems is required by recent technological innovations, which led to the pervasive diffusion of increasingly complex digital systems for the control and supervision of physical systems. The study of hybrid systems is generally more challenging than that of purely discrete or purely continuous systems, because of the interaction between dynamics of different nature. Models for hybrid systems will be introduced in this course and general methods to investigate their properties will be described. Control of hybrid systems will also be addressed by focusing on some case studies from different application contexts. Students attending the course should, at the end of the second module, be able to appreciate the diversity of phenomena that arise in hybrid systems, and understand how concepts that are classical in the theory of discrete event systems, modeled by automata, can coexist with concepts that are classical in the theory of continuous systems, modeled by differential or difference equations, in a unifying framework.

- System Theory - Automatic Control - Fundamentals of statistics and probability - Fundamentals of optimal control

Sutton RS, Barto AG. Reinforcement learning: An introduction. MIT press; 2018.

Classroom attendance is not mandatory.

Project evaluation. The final grade will be based on a project presentation. The project has to be an independent achievement, where the goals and contents are defined by each student in agreement with professors.

Frontal teaching.