Shi Group at TU Delft
Shi Group at TU Delft
Embedded Files
Our research focuses on leveraging data and learning to develop control methods for problems that are high-dimensional or do not have an accurate model from first principles. We are interested in both fundamental theory, methods, and healthcare-related applications:
Application:
Learning and control approaches for human motion digital twin in healthcare
Learning biological networks
Theory & Method:
Learning-based control with performance and safety guarantees
Learning dynamical systems
News
News
Dec. 2025 Changrui Liu (PhD student at Delft) will present his work on learning-based control at CDC 2025, Brazil. The work is titled "On the Regret of Model Predictive Control With Imperfect Inputs" and published in IEEE Control Systems Letters: Link to IEEE
Nov. 2025 Our IEEE TAC paper on safe learning control and control barrier function is now online (early access): Link to IEEE, which was led by Kanghui He, PhD student at Delft.
Oct. 2025: I attended the IROS 2025 conference in Hangzhou, China.
Sep. 2025: Our new preprint "Bang-Ride Optimal Control: Monotonicity, External Positivity, and Fast Battery Charging" is now on ArXiv and is submitted to IEEE TAC, a wonderful collaboration with colleagues from MIT, Stanford, and TU Darmstadt!
Sep. 2025: Our new IEEE TAC paper "Suboptimality analysis of receding horizon quadratic control with unknown linear systems and its applications in learning-based control" is now online!
Aug. 2025: Kelvin Toonen (MSc from TU/e) has started as a PhD student in my group, working on learning and control for human digital twin
Aug. 2025: I have started as an Assistant Professor at TU Delft!
Mar. 2024: I have joined MIT as a PostDoc, working on optimal control for large-scale nonlinear systems
Page updated
Google Sites
Report abuse