Ivan Markovsky

Ivan Markovsky

Centre Internacional de Mètodes Numèrics a Enginyeria

Engineering Sciences

Ivan Markovsky is an ICREA research professor at the International Centre for Numerical Methods in Engineering (CIMNE). His background is in electrical engineering with a PhD from the Katholieke Universiteit Leuven, Belgium, in 2005. From 2006 to 2012 he lectured and did research at the School of Electronics and Computer Science of the University of Southampton, U.K. He continued his academic career with a 10-years research professor mandate 2012-2022 at the Electrical Engineering Department of the Vrije Universiteit Brussel, Belgium. After a 3-month visiting professor position at the Automatic Control Laboratory (IfA) of the ETH Zurich, Switzerland, he moved in 2023 to CIMNE. His expertise is in structured low-rank approximation, system identification, and data-driven control. On these topics he has 110 peer-reviewed papers, 11 book chapters, and 2 monographs.
In 2011, he was awarded an ERC starting grant on the topic of structured low-rank approximation.

Research interests

The objective of my research is unsupervised data-driven analysis and design of dynamical systems. The classical paradigm splits the problem into model identification and model-based design. In general, there is no separation principle for modeling and design, so that the two-stage approach may be suboptimal. I am investigating an alternative direct data-driven paradigm that combines modeling and design into one joint problem. In 2010, I proposed a solution approach for data-driven design based on structured low-rank approximation (ERC starting grant). More recently, I investigated convex relaxation, subspace, and regularization methods. Current topics of interest are data-driven methods for nonlinear, time-varying, and distributed systems. Besides data-driven design, I am interested in methods for teaching and learning that are effective in training critical thinking and creativity. I am an advocate of the open peer review as an alternative to the traditional closed review system.

Selected publications

Markovsky I, Huang L & Dörfler F 2023, ‘Data-driven control based on behavioral approach: From theory to applications in power systems’, IEEE Control Systems Magazine, 43, 28-68.

Markovsky I & Dorfler F 2023, ‘Identifiability in the Behavioral Setting‘, Ieee Transactions On Automatic Control, 68, 3, 1667 – 1677.

Markovsky I, Prieto-Araujo E & Doerfler F 2023, ‘On the persistency of excitation’, Automatica, 147, 110657.

– Dorfler F, Coulson J & Markovsky I 2023, ‘Bridging Direct and Indirect Data-Driven Control Formulations via Regularizations and Relaxations‘, Ieee Transactions On Automatic Control, 68, 2, 883 – 897.

Markovsky I 2023, ‘Data-Driven Simulation of Generalized Bilinear Systems via Linear Time-Invariant Embedding‘, Ieee Transactions On Automatic Control, 68, 2, 1101 – 1106.

– Fazzi A & Markovsky I 2023 ‘Distance problems in the behavioral setting‘, European journal of control, 74.

Selected research activities

– Invited talk “Behavioral approach to system identification and data-driven signal processing” given at the Seminar Series on Optimization, Learning and Control, Laboratoire d’Automatique, EPFL, Switzerland, 10 March.
– Invited talk “Direct data-driven analysis, signal processing, and control” given at the Kolloquium Technische Kybernetik, University of Stuttgart, Germany, 6 June.
– Research visit for collaboration with Florian Dörfler and his group on a joint project “From model-based to data-driven design: Signal processing and control of noisy nonlinear systems”, IfA lab of ETH-Zurich, Switzerland, 1-31 August.
– Invited talk “Optimization problems in data-driven control” given at the Optimization days workshop, University of Southampton, 28 September.
– Talk “System theory without state-space and transfer functions? Yes, it’s possible!” given at STADIUS group, K.U.Leuven, Belgium, 4 December.