Name
Abstract | ||
|---|---|---|
The paper proposes to reduce complexity of explicit MPC feedback laws by representing regions over which the law is defined as (possibly non-convex) polygons. Each polygon is then represented only by its boundaries, which reduces the memory footprint of the feedback law. Even though significant amount of memory can be saved this way, the price to be paid is increased computational load associated by performing point location tasks in non-convex objects. Therefore we propose to devise inner and outer convex approximations of non-convex polygons to reduce the computational requirements. Such approximations allow to perform point location more effectively, leading to a reduction of the required on-line computational effort. Several ways to design suitable approximations are presented and efficacy of the proposed procedure is evaluated. |
Year | DOI | Venue |
|---|---|---|
2013 | 10.1109/CDC.2013.6760905 | Decision and Control |
Keywords | Field | DocType |
approximation theory,computational complexity,concave programming,convex programming,feedback,geometry,predictive control,complexity reduction,computational load,computational requirement reduction,explicit MPC feedback,explicit model predictive control,inner convex approximations,memory footprint reduction,nonconvex polygons,outer convex approximations,point location tasks,polygonic representation,required online computational effort | Polygon,Mathematical optimization,Point location,Computer science,Control theory,Model predictive control,Regular polygon,Memory footprint,Convex optimization,Computational resource,Computational complexity theory | Conference |
ISSN | ISBN | Citations |
0743-1546 | 978-1-4673-5714-2 | 1 |
PageRank | References | Authors |
0.36 | 3 | 4 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Juraj Oravec | 1 | 9 | 3.62 |
| Blazek, S. | 2 | 1 | 0.69 |
| Michal Kvasnica | 3 | 124 | 17.65 |
| S. Cairano | 4 | 249 | 26.23 |