One PhD position on “Data-driven Nonlinear Control of High Precision Robotic Systems” is available.
This project aims to design a data-driven position-dependent controller for high-precision robotic systems, to reduce engineering costs and improve adaptability. The approach employs the Koopman operator to compute a Linear Parameter Varying (LPV) controller and uses Integral Quadratic Constraints (IQC) for robust performance. Validation will be conducted on an industrial high-precision robotic system, with a learning-based module enhancing adaptability to various loads while maintaining robustness. This approach builds on previous work in precision motion systems, demonstrating potential for improving tracking performance through accurate learning-based models and nonlinear optimization.
The project is funded by NCCR Automation and co-supervised by Prof. Alisa Rupenyan from ZHAW.
The interested candidates that have already been accepted in the doctoral school of EPFL can contact Prof. Alireza Karimi .