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.

  

Extension of FORCES solver for the use of programmable logic

Author(s):

S. Merkli
Conference/Journal:

Semester Thesis, FS13 (10285)
Abstract:

The central drawback in Model Predictive Control is its computational complexity. While ecient solvers are being developed in software, this thesis is a study about how emerging system{on{ chip hardware can be used to perform the required mathematical operations faster and more eciently. The FORCES solver by Alexander Domahidi is analyzed for its time usage and the O(n3) operations are rewritten to be more amenable to implementation in programmable logic. A few operations are implemented and analyzed for their timing improvements. A signi cant speedup is found in the implemented versions of the operations, demonstrating the potential of moving more operations into logic. Due to slow data transfer between logic and processor, outsourcing single operations into logic and calculating the rest on a processor is found to be of little bene t. The possibilities of chaining operations in logic are explored and suggestions are made as to how this could be implemented.

Year:

2013
Type of Publication:

(13)Semester/Bachelor Thesis
Supervisor:

A. Domahidi

No Files for download available.
% Autogenerated BibTeX entry
@PhdThesis { Xxx:2013:IFA_4463
}
Permanent link