Matthieu Barreau

Matthieu Barreau is an Assistant Professor within the Division of Decision and Control Systems at KTH Royal Institute of Technology, Stockholm, Sweden, since September 2023. His academic journey includes a PhD in Control Systems from LAAS-CNRS, Toulouse (2019), a Master's degree in Space Engineering from KTH (2016), and an Engineering degree in Aeronautical Engineering from ISAE-ENSICA, Toulouse (2016).

Education:

• PhD in Control Systems, LAAS-CNRS, Toulouse (2019)
• Master's in Space Engineering, KTH (2016)
• Engineering degree in Aeronautical Engineering, ISAE-ENSICA, Toulouse (2016)

Dr. Barreau's research focuses on the intersection of system theory and machine learning, particularly on physics-informed neural networks (PINNs) applied to traffic systems, time-delay systems, and infinite-dimensional systems. His methodology addresses three key questions: what can be expected from data given measurement constraints, how to effectively train PINNs under specific constraints, and which tools are needed to certify the quality of trained models. His work combines robust control theory (Lyapunov functions, Integral Quadratic Constraints) with modern machine learning techniques, creating a bridge between traditional control theory and data-driven approaches.

His recent publications demonstrate a clear trend toward applying PINNs to traffic flow modeling, power system monitoring, and stability analysis of complex dynamical systems. The research spans theoretical foundations of stability analysis for infinite-dimensional systems to practical applications in traffic control and power transformer monitoring. A significant portion of his work addresses the challenge of training physics-informed neural networks under constrained optimization frameworks, developing methods to ensure stability and performance guarantees.

Scientific Awards:

• Best French Ph.D. thesis award from GdR MACS and Club EEA in 2020

Dr. Barreau actively supervises master's students for internships and thesis projects, though specific students aren't listed in the available information. His research is funded by multiple prestigious sources including the Human Horizon project Ultimate, the Swedish Foundation for Strategic Research, the Swedish Research Council, and Knut and Alice Wallenberg Foundation. His international collaborations include researchers from The University of Sheffield, Gipsa-Lab, and other institutions across Europe and the US. He leads projects on traffic applications under the supervision of Karl-Henrik Johansson and collaborates on stability analysis of infinite-dimensional systems with Alexandre Seuret, Frederic Gouaisbaut, and Carsten Scherer.

Dr. Barreau has developed notable open-source contributions including the tf2-bfgs package for implementing BFGS optimization in TensorFlow 2, demonstrating his commitment to making research tools accessible to the broader community.