Murat Arcak Professor Office: 569 Cory Hall Email: arcak@*

Research

Our group develops scalable control design and verification methods for applications where the system size and complexity are beyond the reach of traditional methods. We particularly address systems with many interconnected components, nonlinear dynamics, and learning and decision making capabilities. Our broad approach is to obtain system-level guarantees from appropriate abstractions of the components with compositional, hierarchical and learning-based procedures, and to exploit structural system properties for scalability. We uncover such properties in applications and generalize them to broader classes of systems. Current research topics include control of autonomous systems, networks and multi-agent systems. Please follow the link to Research Areas on the left for further descriptions of these topics.

Course Materials

• Nonlinear Systems (EE C222/ME C237), Spring 2018 (website with class notes)

• Dissipation Inequalities and Quadratic Constraints for Control, Optimization, and Learning - short course taught with Peter Seiler (U. Michigan) as part of the European Embedded Control Institute International Graduate School on Control (May 2024) (website with slides and videos)

• Signals and Systems (EE 120), Fall 2019 (website with class notes)

• Designing Information Devices and Systems II (EE 16B), Spring 2017 (notes)

Affiliations

• CITRIS People and Robots

• Berkeley Institute of Transportation Studies

• Center for the Theoretical Foundations of Learning, Inference, Information, Intelligence, Mathematics and Microeconomics at Berkeley (CLIMB)

* .berkeley.edu