Real-Time Hybrid System Identification and Learning-based Control
Speaker: Christos Mavridis
DEIB - Conference Room "E. Gatti" (Bld. 20)
June 18th 2025 | 2.30 pm
Contact: Prof. Maria Prandini
DEIB - Conference Room "E. Gatti" (Bld. 20)
June 18th 2025 | 2.30 pm
Contact: Prof. Maria Prandini
Sommario
On June 18th, 2025 at 2.30 pm the seminar titled "Real-Time Hybrid System Identification and Learning-based Control" will take place at DEIB "Emilio Gatti" Conference Room (Building 20).
The growing integration of intelligent autonomous systems in modern society necessitates new advancements in cyber-physical systems that can learn, adapt, and reason. Towards this direction, we will investigate the frontiers of real-time hybrid system identification, where discrete and continuous dynamics are interlaced. In particular, we will explore the role of homotopy optimization in learning, introducing the Online Deterministic Annealing (ODA) approach as a gradient-free stochastic optimization method to identify the unknown modes of a hybrid system in real-time. We will study the properties of robustness and interpretability and the significance of controlling the performance-complexity trade-off through an intuitive bifurcation phenomenon. Finally, we will discuss a wide range of applications, from learning-based control and communication-aware motion planning to cyber-physical systems security.
The growing integration of intelligent autonomous systems in modern society necessitates new advancements in cyber-physical systems that can learn, adapt, and reason. Towards this direction, we will investigate the frontiers of real-time hybrid system identification, where discrete and continuous dynamics are interlaced. In particular, we will explore the role of homotopy optimization in learning, introducing the Online Deterministic Annealing (ODA) approach as a gradient-free stochastic optimization method to identify the unknown modes of a hybrid system in real-time. We will study the properties of robustness and interpretability and the significance of controlling the performance-complexity trade-off through an intuitive bifurcation phenomenon. Finally, we will discuss a wide range of applications, from learning-based control and communication-aware motion planning to cyber-physical systems security.
Biografia
Christos Mavridis received his Diploma in electrical and computer engineering from the National Technical University of Athens, Greece, in 2017, and the M.S. and Ph.D. degrees in electrical and computer engineering at the University of Maryland, College Park, MD, in 2021. His research interests include stochastic optimization, systems and control theory, learning theory, multi-agent systems, and robotics.
He is currently a postdoc at KTH Royal Institute of Technology, Stockholm, and has been affiliated as a research scientist with the Nokia Bell Labs, NJ, the Xerox Palo Alto Research Center (PARC), CA, and Ericsson AB, Stockholm.
Dr. Mavridis is an IEEE member, and a member of IEEE/CSS Technical Committee on Security and Privacy. He has received the A. James Clark School of Engineering Distinguished Graduate Fellowship and the Ann G. Wylie Dissertation Fellowship in 2017 and 2021, respectively. He has been a finalist in the Qualcomm Innovation Fellowship US, San Diego, CA, 2018, and he has received the Best Student Paper Award in the IEEE International Conference on Intelligent Transportation Systems (ITSC).
He is currently a postdoc at KTH Royal Institute of Technology, Stockholm, and has been affiliated as a research scientist with the Nokia Bell Labs, NJ, the Xerox Palo Alto Research Center (PARC), CA, and Ericsson AB, Stockholm.
Dr. Mavridis is an IEEE member, and a member of IEEE/CSS Technical Committee on Security and Privacy. He has received the A. James Clark School of Engineering Distinguished Graduate Fellowship and the Ann G. Wylie Dissertation Fellowship in 2017 and 2021, respectively. He has been a finalist in the Qualcomm Innovation Fellowship US, San Diego, CA, 2018, and he has received the Best Student Paper Award in the IEEE International Conference on Intelligent Transportation Systems (ITSC).
