Research projects

My research focuses on learning, analyzing, and controlling cyber-physical systems, with an emphasis on safety and correctness guarantees. The table below connects each topic with selected publications; see the full publication list for more details.

Current research topics and selected publications
Research topicSelected references

Data-driven analysis and control

Learning safety certificates and controllers directly from data when an accurate system model is unavailable. I study Lyapunov and barrier functions, feedback control, and model predictive control with probabilistic guarantees.

Formal verification

Certifying the safety and correctness of nonlinear and hybrid systems for safety-critical applications. My methods combine optimization, counterexample-guided synthesis, neural networks, and SMT solvers.

System identification

Learning models of cyber-physical systems from observations, especially when first-principles modeling is difficult. I develop algorithms for identifying switched and piecewise affine dynamics.

CC BY-SA 4.0 Guillaume Berger. Website built with Hugo.