|
SUBSPACE IDENTIFICATION - REDUCING UNCERTAINTY ON THE STOCHASTIC PART
Torben Knudsen
Aalborg University, Institute of Electronic Systems, Department of
Control Engineering, Fredrik Bajers Vej 7C, DK-9220 Aalborg Ø,
Denmark, e-mail, tk@control.auc.dk
Subspace identification algorithms are user friendly, numerical fast and stable and they provide a good consistent estimate of the deterministic part of a system. The weak point is the stochastic part. The uncertainty on this part is discussed below and methods to reduce it is derived.
Keywords: Parameter estimation; linear multivariable systems; subspace methods; state space models; bias; stochastic part; least-squares methods
Session slot T-Mo-M02: State Space Identification and Subspace Methods/Area code 3a : Modelling, Identification and Signal Processing
 | 