MPC Toolbox with GPU Accelerated Optimization Algorithms

Nicolai Fog Gade-Nielsen, John Bagterp Jørgensen, Bernd Dammann

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

1177 Downloads (Pure)

Abstract

The introduction of Graphical Processing Units (GPUs) in scientific computing has shown great promise in many different fields. While GPUs are capable of very high floating point performance and memory bandwidth, its massively parallel architecture requires algorithms to be reimplemented to suit the different architecture. Interior point method can be used to solve convex optimization problems. These problems often arise in fields such as in Model Predictive Control (MPC), which may have real-time requirements for the solution time. This paper presents a case study in which we utilize GPUs for a Linear Programming Interior Point Method to solve a test case where a series of power plants must be controlled to minimize the cost of power production. We demonstrate that using GPUs for solving MPC problems can provide a speedup in solution time.
Original languageEnglish
Title of host publicationThe 10th European Workshop on Advanced Control and Diagnosis (ACD 2012)
Number of pages6
PublisherTechnical University of Denmark
Publication date2012
Publication statusPublished - 2012
Event10th European Workshop on Advanced Control and Diagnosis - Technical University of Denmark, Kgs. Lyngby, Denmark
Duration: 8 Nov 20129 Nov 2012
http://indico.conferences.dtu.dk/conferenceDisplay.py?confId=108

Conference

Conference10th European Workshop on Advanced Control and Diagnosis
LocationTechnical University of Denmark
CountryDenmark
CityKgs. Lyngby
Period08/11/201209/11/2012
Internet address

Keywords

  • Linear programming
  • Interior Point Methods
  • Model predictive control
  • Graphical Processing Unit

Cite this

Gade-Nielsen, N. F., Jørgensen, J. B., & Dammann, B. (2012). MPC Toolbox with GPU Accelerated Optimization Algorithms. In The 10th European Workshop on Advanced Control and Diagnosis (ACD 2012) Technical University of Denmark.