PhD Candidate in Engineering Science

Department of Electrical Engineering (ESAT), STADIUS division

KU Leuven

Renzi Wang received her bachelor’s degree in mechatronics with a focus on robotics from Harbin Institute of Technology, China, in 2018, and her master’s degree in mechatronics with a focus on control theory from Karlsruhe Institute of Technology, Germany, in 2021. During her master’s study, she took one exchange semester at EPFL, Switzerland, and completed her master thesis with a topic on extending a data-enabled predictive control method on nonlinear systems there.

Project description

Autonomous navigation in uncertain environments, where vehicles (autonomous cars, robots, drones) must interact with other vehicles or other users (e.g., pedestrians) was identified as a highly challenging control task. Thus, Renzi developed MPC strategies that took into account interaction and uncertainty. Traditional means of handling uncertainty such as robust or stochastic approaches were demonstrated to be either too conservative or too risk-prone. Moreover, existing MPC methodologies typically assumed that the uncertain behavior of surrounding users was completely exogenous, thus failing to take interaction into account. To explicitly account for interactions, Renzi developed stochastic models for surrounding users whose probabilistic structure depended on the states of other users. The parameter models were learned through historical trajectories using machine learning. The resulting NMPC formulations led to optimal control problems that were more complex and of larger scale than what state-of-the-art embedded solvers could handle. Renzi’s project successfully developed optimization algorithms for interaction-aware MPC for autonomous navigation in uncertain environments, including a novel EM++ parameter learning framework for stochastic switching systems and risk-sensitive MPC approaches with sequential convexification algorithms.

Publications

1.

Wang, Renzi; Schuurmans, Mathijs; Patrinos, Panagiotis

Risk-Sensitive Model Predictive Control for Interaction-Aware Planning--A Sequential Convexification Algorithm Working paper

2025.

Abstract | Links | BibTeX

2.

Wang, Renzi; Acerbo, Flavia Sofia; Son, Tong Duy; Patrinos, Panagiotis

Imitation Learning from Observations: An Autoregressive Mixture of Experts Approach Working paper

2024.

Abstract | Links | BibTeX

3.

Wang, Renzi; Bodard, Alexander; Schuurmans, Mathijs; Patrinos, Panagiotis

EM++: A parameter learning framework for stochastic switching systems Working paper

2024.

Abstract | Links | BibTeX

4.

Wang, Renzi; Schuurmans, Mathijs; Patrinos, Panagiotis

Interaction-aware Model Predictive Control for Autonomous Driving Proceedings Article

In: 2023 European Control Conference (ECC), pp. 1-6, IEEE, Bucharest, Romania, 2023, ISBN: 978-3-907144-08-4.

Abstract | Links | BibTeX