Master Thesis

Model Predictive Control for Robotic Cloth Manipulation

Work default illustration

Information

  • If you are interested in the proposal, please contact with the supervisors.

Description

This project aims at using Model Predictive Control [1] an a cloth model being developed at our institute in order to manipulate cloth with robotic arms.

Model predictive Control uses a system model in order to perform a control action while keeping in mind a prediction of the system's behavior. We will use a new formulation for cloth modelling and non-linear model predictive control to make a robot move a certain piece of cloth, following our desired motion.

This project aims at students looking for challenges in the emerging field of MPC. Some skills recommended for the candidate are:
- Some background in programming (matlab and/or C++)
- Control background
- Non-linear optimization
- Good level of algebra and calculus

The project also aims at publishing the results of this project in a conference.

[1] Lars Grüne, Jürgen Pannek. "Nonlinear Model Predictive Control". Springer Communications and Control Engineering series. Springer, 2017.