Formal control synthesis are correct-by-construction approaches to automatically create controllers for autonomous systems. These controllers aim to accomplish complex tasks specified using temporal logic. Synthesis is challenging when there is uncertainty in the system, such as the presence of black-box components or incomplete dynamics knowledge, or in the environment. Machine-learning regression can construct models of these uncertainties that can be used to reason about the future.
Our goal is to combine machine learning regression with formal control synthesis approaches to address these uncertainties. Mature learning approaches, such as Gaussian process regression, have a strong analytical results that quantify the regression errors. Our current approach involves constructing abstractions of the system using the learned model, and then use model-based techniques to synthesize a controller. Ongoing challanges we are investigating include:
- Scalable computational methods for abstraction
- Utilizing other approaches for GP regression and machine learning
- Implementation on robotic systems and vehicles
If you are interested in learning more, please reach out to John Jackson at john.m.jackson@colorado.edu or checkout some of our recent related publications.
2024
-
R. Reed, H. Schaub, and M. Lahijanian, “Shielded Deep Reinforcement Learning for Complex Spacecraft Specifications,” in American Control Conference (ACC), Toronto, Canada, 2024. (accepted)
-
F. B. Mathiesen, M. Lahijanian, and L. Laurenti, “IntervalMDP.jl: Accelerated Value Iteration for Interval Markov Decision Processes,” in IFAC Conference on Analysis and Design of Hybrid Systems (ADHS), 2024. (accepted)
2023
-
J. Skovbekk, L. Laurenti, E. Frew, and M. Lahijanian, “Formal Abstraction of General Stochastic Systems via Noise Partitioning,” IEEE Control Systems Letters, pp. 1–1, Dec. 2023.
-
R. Reed, L. Laurenti, and M. Lahijanian, “Promises of Deep Kernel Learning for Control Synthesis,” IEEE Control Systems Letters, pp. 1–1, Dec. 2023.
-
S. Adams, A. Patane, M. Lahijanian, and L. Laurenti, “BNN-DP: Robustness Certification of Bayesian Neural Networks via Dynamic Programming,” in International Conference on Machine Learning (ICML), Honolulu, HI, USA, 2023, pp. 133–151.
-
G. Delimpaltadakis, M. Lahijanian, M. Mazo Jr., and L. Laurenti, “Interval Markov Decision Processes with Continuous Action-Spaces,” in International Conference on Hybrid Systems: Computation and Control (HSCC), San Antonio, TX, USA, 2023.
2022
-
R. Mazouz, K. Muvvala, A. Ratheesh Babu, L. Laurenti, and M. Lahijanian, “Safety Guarantees for Neural Network Dynamic Systems via Stochastic Barrier Functions,” in Advances in Neural Information Processing Systems (NeurIPS), New Orleans, Louisiana, USA, 2022, vol. 35, pp. 9672–9686.
-
S. A. Adams, M. Lahijanian, and L. Laurenti, “Formal Control Synthesis for Stochastic Neural Network Dynamic Models,” IEEE Control Systems Letters (L-CSS), 2022.
2021
-
J. Jackson, L. Laurenti, E. Frew, and M. Lahijanian, “Synergistic Offline-Online Control Synthesis via Local Gaussian Process Regression,” in IEEE Conference on Decision and Control (CDC), 2021.
-
J. Jackson, L. Laurenti, E. Frew, and M. Lahijanian, “Towards Safe, Abstraction-based Online Learning and Synthesis for Unknown Systems,” in Workshop on Integrating Planning and Learning, 2021.
-
J. Jackson, L. Laurenti, E. Frew, and M. Lahijanian, “Strategy synthesis for partially-known switched stochastic systems,” in Conference on Hybrid Systems: Computation and Control (HSCC), 2021, pp. 1–11.
2020
-
J. Jackson, L. Laurenti, E. Frew, and M. Lahijanian, “Towards Data-driven Verification of Unknown Dynamical Systems,” in Workshop on Robust Autonomy: Tools for Safety in Real-World Uncertain Environments, 2020.
-
J. Jackson, L. Laurenti, E. Frew, and M. Lahijanian, “Safety verification of unknown dynamical systems via gaussian process regression,” in 2020 IEEE 59th Conference on Decision and Control (CDC), 2020.