RDP is short for "Robust Dynamic Programming". RDP is a collection of Matlab functions that implement a method for robust dynamic programming for linearly constrained polytopic systems with piecewise linear cost, described in the articles listed below. The software was developed at the Interdisciplinary Center for Scientific Computing (IWR) in Heidelberg, Germany, in the summers 2003 and 2004 by Jakob Björnberg and Moritz Diehl. J. Björnberg visited IWR these summers to work under the supervision of M. Diehl.
To use rdp, you need cddmex by Mato Baotic and Torrisi Fabio, which is a Matlab interface for the program cdd by Komei Fukuda. The easiest way to start using rdp is to download the pre-compiled binaries for cddmex that are available at the cddmex homepage.
Download rdp:
Manual for rdp (included in the rdp package):
Articles:
-
Moritz Diehl and Jakob Björnberg. Robust dynamic programming for min-max model predictive control of constrained uncertain systems. IEEE Transactions on Automatic Control, Vol 49, No 12, pp. 2253-2257. December 2004. Preprint as PS- or PDF- file. (BibTeX entry)
- Jakob Björnberg and Moritz Diehl. Approximate robust dynamic programming and robustly stable MPC. Automatica, 42(5), pp. 777-782, 2006 (BibTeX entry)
Keywords: dynamic programming, receding horizon control (RHC), min-max model predictive control (MPC), robustness.
