The Parallel Library in Control: Model Reduction (PLiCMR) package contains subroutines for model reduction of linear dynamical systems in state-space form on parallel distributed computers. Currently it contains routines for balanced truncation, singular perturbation approximation, optimal Hankel norm approximation, and stochastic balancing for stable systems using dense matrix algebra.
The parallelization model is based on the approach used in ScaLAPACK. Numerical computations are based on ScaLAPACK as far as possible and communication is performed using the Basic Linear Algebra Communication Subroutines (BLACS) on top of a communication library such as PVM or MPI.
The documentation of the library is available on-line in HTML format.
For more information on PLiCMR, consult
-
Peter Benner, Enrique S. Quintana-Ortí, and Gregorio Quintana-Ortí
PSLICOT Routines for Model Reduction of Stable Large-Scale Systems
Proceedings of the Third NICONET Workshop on Numerical Software in Control Engineering,
Louvain-la-Neuve, Belgium, January 19, 2001, pp. 39-44.
-
Peter Benner, Enrique S. Quintana-Ortí, and Gregorio Quintana-Ortí
State-Space Truncation Methods for Parallel Model Reduction of Large-Scale Systems
Parallel Computing, Vol. 29 (special issue "Parallel and Distributed Scientific and Engineering Computing"), pp. 1701-1722, 2003.
which describes those routines of PLiCMR that have been integrated into the parallelized version of SLICOT, called PSLICOT.