Title :
Multi-mode system identification
Author :
Verriest, Erik I. ; De Moor, Bart
Author_Institution :
Sch. of ECE, Georgia Inst. of Technol., Atlanta, GA, USA
fDate :
Aug. 31 1999-Sept. 3 1999
Abstract :
The subspace identification method for time-invariant linear systems is extended to the class of multi-mode or hybrid systems. Under an excitation criterion, the parameters in all modes are simultaneously estimated, and combined with a detection scheme for the mode switching. In turn, parameters of the mode switching mechanism, whether deterministic or Markovian, can be estimated as well. The method is readily applicable to fault detection and identification.
Keywords :
continuous systems; discrete systems; identification; linear systems; hybrid systems; mode switching mechanism; multimode system identification; subspace identification method; time-invariant linear systems; Data models; Fault detection; Mathematical model; Silicon; Stacking; Switches; Identification; Multi-Mode Systems; Subspace Methods; Time-Varying Systems;
Conference_Titel :
Control Conference (ECC), 1999 European
Conference_Location :
Karlsruhe
Print_ISBN :
978-3-9524173-5-5
