Title :
Balanced canonical forms for system identification
Author :
Bauer, D. ; Deistler, M.
Author_Institution :
Inst. fur Econometrics, Operations Res. & Syst. Theory, Tech. Univ. Wien, Austria
Abstract :
An algorithm for the identification of linear systems based on balanced parametrizations and optimization of a (e.g. likelihood type) criterion function is proposed and analyzed. The topological properties of the parametrization, which are relevant for the consistency of the parameters, are analyzed. The algorithm contains both model selection steps and estimation of real valued parameters. One of the appealing features of the algorithm is that for given order at most 3 optimization steps have to be performed
Keywords :
controllability; linear systems; parameter estimation; state-space methods; topology; balanced canonical forms; balanced parametrizations; criterion function; linear systems; model selection; system identification; Algorithm design and analysis; Covariance matrix; Ear; Econometrics; Linear systems; MIMO; Operations research; State-space methods; System identification; Transfer functions;
Conference_Titel :
Decision and Control, 1997., Proceedings of the 36th IEEE Conference on
Conference_Location :
San Diego, CA
Print_ISBN :
0-7803-4187-2
DOI :
10.1109/CDC.1997.650696
