Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.


Solution of Large Scale MPC Problems on a GPU and multi-core CPU


Stefan Nesic

Master Thesis, HS12 (10175)

In nowadays applications, constant challenges occur in solving finite horizon optimal control problems in real time, especially those that have large scale number of state and control variables. In this work, we intent to present and elaborate on appropriate parallelization techniques that can be used to successfully implement two first-order optimization methods in parallel manner for solving finite horizon optimal control problems. The particular methods under investigation are the alternating direction method of multipliers (ADMM) and the fast gradient method (FGM) that we use for application of model predictive control (MPC) of large systems. Parallelization of optimization methods is only applicable to computing architectures that support it, such as graphical processing units (GPUs) and multi-core CPUs that we discuss within this thesis. Computational experiments on differently structured optimal control problems of various sizes showed that GPU implementations can yield a significant speedup compared to state-of-the-art CPU based solvers such as CPLEX.


Type of Publication:

(12)Diploma/Master Thesis

C. Conte

No Files for download available.
% Autogenerated BibTeX entry
@PhdThesis { Xxx:2013:IFA_4441
Permanent link