Consistency and asymptotic normality of some subspace algorithms for systems without observed inputs

DOI10.1016/S0005-1098(99)00031-XMaRDI QIDQ1301438zbMATH OpenOpenAlexFDO

Authors Dietmar Bauer, Manfred Deistler, Wolfgang Scherrer

Publication date 2 September 1999

Published in Automatica (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1016/s0005-1098(99)00031-x

zbMATH Keywords

asymptotic analysisasymptotic normalityconsistencydiscrete time systemsidentificationlinear multivariable systemssubspace algorithms

Mathematics Subject Classification ID

System identification (93B30) Identification in stochastic control theory (93E12)

Review text

For a class of subspace algorithms, in the case where no observed inputs are included, the asymptotic properties of the estimates of the system matrices are discussed. One proves the asymptotic normality of subspace estimates and that the variance may be calculated. A new consistency result for the system matrix is presented.

Recommendations

• Analysis of the asymptotic properties of the MOESP type of subspace algorithms
• On consistency of subspace methods for system identification
• Asymptotic properties of subspace estimators
• Consistency and relative efficiency of subspace methods

Cited in

(23)

• Estimating ARMAX systems for multivariate time series using the state approach to subspace algorithms
• On consistency of subspace methods for system identification
• The asymptotic variance of subspace estimates.
• The relation of the CCA subspace method to a balanced reduction of an autoregressive model.
• Numerical conditioning and asymptotic variance of subspace estimates
• Some facts about the choice of the weighting matrices in Larimore type of subspace algorithms
• Asymptotic properties of subspace estimators
• Estimating cointegrated systems using subspace algorithms
• Consistency and relative efficiency of subspace methods
• Analysis of the asymptotic properties of the MOESP type of subspace algorithms
• Model specification and selection for multivariate time series
• An improved bias-compensation approach for errors-in-variables model identification
• Subspace-based fault detection robust to changes in the noise covariances
• USING SUBSPACE METHODS FOR ESTIMATING ARMA MODELS FOR MULTIVARIATE TIME SERIES WITH CONDITIONALLY HETEROSKEDASTIC INNOVATIONS
• System identification methods for (operational) modal analysis: review and comparison
• A Note on the Specification and Estimation of ARMAX Systems
• Comparing the CCA Subspace Method to Pseudo Maximum Likelihood Methods in the case of No Exogenous Inputs
• ESTIMATING LINEAR DYNAMICAL SYSTEMS USING SUBSPACE METHODS
• Subspace identification for non-linear systems with measured-input non-linearities
• Bootstrap-based estimates of uncertainty in subspace identification methods
• Non-asymptotic error analysis of subspace identification for deterministic systems
• Learning linearized models from nonlinear systems under initialization constraints with finite data
• Finite sample analysis of subspace identification for autonomous stochastic systems

This page was built for publication: Consistency and asymptotic normality of some subspace algorithms for systems without observed inputs

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1301438)