Research
My group studies learning–control methods with formal guarantees. Two themes anchor our recent work: Networked Control with Safety & Stability and System Theory for Algorithm Design.
1) Networked Control with Safety and Stability Guarantees
Large-scale networked systems such as autonomous aerial/ground fleets, power grids—require many agents to coordinate sensing, actuation, and communication in real time. We design learning-based control frameworks that scale to these environments while preserving rigorous safety and stability guarantees, without unnecessary conservatism.
Slide decks
Representative publications
- Sparsity Invariance for Convex Design of Distributed Controllers , IEEE Transactions on Control of Network Systems, 2020.
- Learning the globally optimal distributed LQ regulator , L4DC 2020.
- Distributed neural network control with dependability guarantees: a compositional port-Hamiltonian approach , L4DC 2022.
- Learning to boost the performance of stable nonlinear systems , IEEE Open Journal of Control Systems, 2024.
- MAD: A Magnitude And Direction Policy Parametrization for Stability Constrained Reinforcement Learning , CDC 2025.
2) System Theory for Algorithm Design
Many engineering problems involve solving optimization tasks rapidly and reliably. System and control theory offer a principled toolkit to design optimization algorithms with formal guarantees on convergence, speed, and robustness. We develop enhanced optimization and machine learning algorithms by leveraging nonlinear system theory and integrating neural network components.
Slide decks
Representative publications
- Learning to optimize with convergence guarantees using nonlinear system theory , IEEE Control Systems Letters, 2024.
- Learning to optimize with guarantees: a complete characterization of linearly convergent algorithms , 2025 (preprint, with A. Martin and I. Manchester).