I am a postdoctoral researcher at the Department of Electrical and Systems Engineering, University of Pennsylvania.
My research interests include data-driven identification and control, nonparametric systems, and Bayesian methods for safe learning.
In modern control applications such as autonomous driving or human-robot-interaction, the control design and modelling process becomes very time-consuming or even infeasible due to the complexity of the systems. A solution to overcome these issues is provided by data-driven models which require only a minimum of prior knowledge.
However, a general problem of data-driven models is the estimation of the model accuracy and a guarantee for the stability of the control loop. My research focuses on Bayesian models, especially Gaussian Processes (GP) since they provide not only a mean prediction, but also a variance as uncertainty measure of the model. I derive GP based control laws which guarantee safety and high-performance, analyze the control properties of GP models and provide guidelines for kernel selection.
Thomas Beckers is a postdoctoral researcher at the Department of Electrical and Systems Engineering, University of Pennsylvania. He is member of the GRASP Lab and the PRECISE Center. In 2020, he earned his doctorate in Electrical Engineering at the Technical University of Munich (TUM), Germany. He received the B.Sc. and M.Sc. degree in Electrical Engineering in 2010 and 2013, respectively, from the Technical University of Braunschweig, Germany. In 2018, he was a visiting researcher at the University of California, Berkeley. For his doctoral thesis he was awarded with the Rhode & Schwarz Outstanding Dissertation price. His research interests include physics-enhanced learning for modeling and control, nonparametric systems, and formal methods for safe learning.
12/2021 Our paper entitled "Online learning-based trajectory tracking for underactuated vehicles with uncertain dynamics” was accepted for publication at IEEE Control Systems Letters (L-CSS)
11/2021 Our paper entitled "Prediction with Approximated Gaussian Process Dynamical Models" was accepted for publication at IEEE Transaction on Automatic Control
11/2021 I received the Rhode & Schwarz Outstanding Dissertation Award for my work on "Gaussian Process based Modeling and Control with Guarantees"
07/2021 We will organize an invited session about "Gaussian Process based Identification and Control" at CDC 2021. Stay tuned for more information!
07/2021 Both of our papers have been accepted for presentation at CDC 2021
05/2021 Our article entitled "Gaussian Process Based Visual Pursuit Control with Unknown Target Motion Learning in Three Dimensions" was accepted for publication in the SICE Journal of Control, Measurement, and System Integration
05/2021 Submission of our article "Online Learning-based Balancing of Feed-forward and Feedback Control"
03/2021 Submission of our articles "Safe Online Learning-based Formation Control of Multi-Agent Systems with Gaussian Processes" and “Cooperative Visual Pursuit Control with Learning of Position Dependent Target Motion via Gaussian Process”
03/2021 Submission of our article "Safe Trajectory Tracking for Underactuated Vehicles with Partially Unknown Dynamics"
02/2021 Check out my Gaussian Process Regression app
01/2021 Submission of our article "Gaussian Process Based Visual Pursuit Control with Unknown Target Motion Learning in Three Dimensions"
01/2021 I've joined George Pappas' group at University of Pennsylvania as postdoctoral researcher
12/2020 I successfully defend my PhD thesis entitled "Gaussian Process based Modeling and Control with Guarantees"
11/2020 Submission of our new article "The Value of Data in Learning-Based Control for Training Subset Selection"
10/2020 Submission of our new article "Real-time Uncertainty Decomposition for Online Learning Control"