This project, supported by CENIIT, is the core project for Johan Löfberg. It is a general project which aims to investigate a wide range of concepts in the field of MPC and neighbouring topics in optimization.
More information about the project and current status can be found here.
A Swedish project definition can be found here.
This project, supported by CENIIT, and forms the back-bone for Daniel Axehill's research about efficient optimization algorithms for MPC. The main focus so far has been to develop fast and efficient optimization algorithms. In the future, we hope also to gain some more insight how to analyze the behavior of these algorithms already off-line in order predict and guarantee their performance when they are applied in closed-loop on-line.
The central topic in this project is development of control laws for aerospace systems which operate under constraints. A natural candidate to investigate in such a setting is of course MPC. A dedicated project page can be found here
This project is supported by the Swedish Research Council (VR) and is related to the real-time MPC project presented above. In this project, the focus is on utilizing the possibilities for parallel computations in modern computer hardware. This includes setups with multiple cores, CPUs, and computers in a clusters. Also GPUs are of interest. The idea is to investigate how these new tools can be utilized for high-performance computations for hybrid MPC problems. Using these tools, the aim is to be able to be able to solver more challenging problems faster.