dc.contributor.author |
Raath, J. |
|
dc.contributor.author |
du Toit Mouton, H. |
|
dc.contributor.author |
Geyer, T. |
|
dc.date.accessioned |
2017-10-05T08:37:54Z |
|
dc.date.available |
2017-10-05T08:37:54Z |
|
dc.identifier.uri |
http://hdl.handle.net/11462/1197 |
|
dc.description |
Published Conference Proceedings |
en_US |
dc.description.abstract |
This paper considers a model predictive controller with reference tracking that manipulates
the integer switch positions of a power converter. It can be shown that the optimal switch position can
be computed without solving an optimization problem. Specifically, in a new coordinate system, the
optimization problem can be solved offline, leading to a polyhedral partition of the solution space. The
optimal switch position can then be found using a binary search tree. This concept is exemplified for a
three-level single-phase converter with an RL load. |
en_US |
dc.format.extent |
838 612 bytes, 1 file |
|
dc.format.mimetype |
Application/PDF |
|
dc.language.iso |
en_US |
en_US |
dc.publisher |
IEEE Xplore: Control and Modeling for Power Electronics (COMPEL), 2016 IEEE 17th Workshop |
en_US |
dc.subject |
Model Predictive Control |
en_US |
dc.subject |
Vector Quantization |
en_US |
dc.subject |
Delaunay triangulation |
en_US |
dc.subject |
Voronoi diagram |
en_US |
dc.subject |
Binary Search Tree |
en_US |
dc.title |
Integration of inverter constraints in geometrical quantification of the optimal solution to an MPC controller |
en_US |
dc.type |
Presentation |
en_US |