Publications
Data-Driven Abstractions via Binary-Tree Gaussian Processes for Formal Verification
Oliver Schön, Shammakh Naseer, Ben Wooding, Sadegh Soudjani
8th IFAC Conference on Analysis and Design of Hybrid Systems (ADHS 2024).
Formal verification of stochastic systems using Binary-Tree Gaussian Processes (BTGPs) for efficient abstraction. BTGPs naturally partition the state space, simplifying error quantification. This yields formal models that are amenable to model checking.
SySCoRe: Synthesis via Stochastic Coupling Relations
Oliver Schön, Birgit van Huijgevoort, Sofie Haesaert, Sadegh Soudjani
IEEE 61st Conference on Decision and Control (CDC 2022).
Establishing formal guarantees while allowing for epistemic uncertainty is still a great challenge. We investigate ways of relating parametric stochastic systems to their nominal model.
With Alessandro Abate, Henk Blom, Joanna Delicaris, Arnd Hartmanns, Abolfazl Lavaei, Hao Ma, Mathis Niehage, Anne Remke, Stefan Schupp, Lisa Willemsen. 9th International Workshop on Applied Verification of Continuous and Hybrid Systems (ARCH22).
A friendly competition for formal verification and policy synthesis of stochastic models showcasing the performance and limits of state-of-the-art solution approaches.
26th ACM International Conference on Hybrid Systems: Computation and Control (HSCC 2023).
Developed MATLAB toolbox for synthesizing robust controllers for stochastic continuous-state systems to satisfy temporal logic specifications and quantifying the associated formal robustness guarantees.
This paper addresses the problem of data-driven computation of controllers that are correct by design for safety-critical systems and can provably satisfy (complex) functional requirements. With a focus on continuous-space stochastic systems with parametric uncertainty, we propose a two-stage approach that decomposes the problem into a learning stage and a robust formal controller synthesis stage. The first stage utilizes available Bayesian regression results to compute robust credible sets for the true parameters of the system. For the second stage, we introduce methods for systems subject to both stochastic and parametric uncertainties. We provide simulation relations for enabling correct-by-design control refinement that are founded on coupling uncertainties of stochastic systems via sub-probability measures. The presented relations are essential for constructing abstract models that are related to not only one model but to a set of parameterized models. The results are demonstrated on three case studies, including a nonlinear and a high-dimensional system.
26th ACM International Conference on Hybrid Systems: Computation and Control (HSCC 2023).
In this ongoing work, we address data-driven computation of controllers that are correct by design for safety-critical systems and can provably satisfy complex functional requirements. We propose a two-stage approach that decomposes the problem into a data-driven stage and a robust formal controller synthesis stage. The first stage utilizes available Bayesian linear regression methods to compute robust confidence sets for the true parameters of the system. The second stage develops methods for systems subject to both stochastic and parametric uncertainties. We provide simulation relations for enabling control refinement that are founded on coupling uncertainties of stochastic systems via sub-probability measures. Such relations are essential for constructing abstract models that are related to not only one model but to a set of parametric models.
IEEE 62st Conference on Decision and Control (CDC 2023).
We address two limitations of existing approaches for formal synthesis of controllers for networks of stochastic systems satisfying complex temporal specifications. Firstly, whilst existing approaches rely on the stochasticity to be Gaussian, the heterogeneous nature of composed systems typically yields a more complex stochastic behavior. Secondly, exact models of the systems involved are generally not available or difficult to acquire. We design controllers based on parameter uncertainty sets identified from observed data and approximate possibly arbitrary noise distributions using Gaussian mixture models whilst quantifying the incurred stochastic coupling.
Alessandro Abate, Henk Blom, Nathalie Cauchi, Joanna Delicaris, Sofie Haesaert, Birgit van Huijgevoort, Abolfazl Lavaei, Anne Remke, Oliver Schön, Stefan Schupp, Fedor Shmarov, Sadegh Soudjani, Lisa Willemsen and Paolo Zuliani
Open challenges are extension to LTL specifications and scalability.
Multi-Objective Physics-Guided Recurrent Neural Networks for Identifying Non-Autonomous Dynamical Systems
Oliver Schön, Ricarda Samantha-Götte, Julia Timmermann
14th IFAC Workshop on Adaptive and Learning Control Systems (ALCOS 2022).
Learning dynamics models from data whilst utilizing physics-based knowledge for efficient convergence and physically plausible models. Combining a multi-objective approach, an approximate dynamics model, and a recurrent neural network.
Theses
Since their introduction, Physics-guided Neural Networks (PGNN), a novel class of hybrid models, have already been successfully implemented in several domains of application. As a result, both synergetic effects as well as physically sound models were obtained. Within the context of this thesis, for the first time, the potential of PGNNs for the identification of dynamic systems is investigated from a control engineering point of view. Various approaches are identified and first investigations are performed. By combining a recurrent neural network with a physical dynamics model, a Physics-guided Recurrent Neural Network (PGRNN) is constructed. It is demonstrated that the PGRNN generally outperforms a purely data-driven approach by a substantial margin. Further investigations indicate that the quality of the introduced dynamics model is merely of minor importance to the resulting performance benefits. Consecutively, the PGRNN is augmented by a physics-based constraint, inciting energy conserving solutions as well as a function library of nonlinear terms. The latter resulted in the full compensation of prediction error deficiencies due to inaccurate dynamic models.
All Publications on Google Scholar
What I have been working on
Other Projects
Coming soon
Meet My Collaborators
Dr. Sadegh Soudjani
Max Planck Institute SWS, Germany
Dr. Sofie Haesaert
TU Eindhoven, The Netherlands
Dr. Birgit van Huijgevoort
TU Eindhoven, The Netherlands
Zhengang Zhong
Imperial College, United Kingdom