Vector ARMA estimation: A reliable subspace approach

DOI10.1109/78.847793MaRDI QIDQ2734358zbMATH OpenOpenAlexFDO

Authors Jørge Mari, Petre Stoica, T. McKelvey

Publication date 13 December 2001

Published in IEEE Transactions on Signal Processing (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1109/78.847793

zbMATH Keywords

autoregressive moving-averagecovariance fittinglinear matrix inequalitypolynomial factorizationsSchur stabilitysemidefinite programmingsubspace methodssystem identificationvector-valued linear stochastic systems

Mathematics Subject Classification ID

Identification in stochastic control theory (93E12) Least squares and related methods for stochastic control systems (93E24)

Review text

The authors propose a subspace approach to parameter estimation for finite-dimensional multivariate linear stochastic systems. This method has the merit that it guarantees the production of valid models approximating the true underlying system in a computational time of a polynomial order in the system dimension. This method is based on a combination of certain stochastic subspace identification techniques with matrix Schur restabilizing procedures and multivariate covariance fitting. Numerical examples are presented to demonstrate the performance of the method.

Recommendations

• Estimating ARMAX systems for multivariate time series using the state approach to subspace algorithms
• scientific article; zbMATH DE number 3913497
• scientific article; zbMATH DE number 4132059
• USING SUBSPACE METHODS FOR ESTIMATING ARMA MODELS FOR MULTIVARIATE TIME SERIES WITH CONDITIONALLY HETEROSKEDASTIC INNOVATIONS
• ESTIMATING LINEAR DYNAMICAL SYSTEMS USING SUBSPACE METHODS

Cited in

(11)

• Combining evolutionary and stochastic gradient techniques for system identification
• Approximate predictor and filter for partially observed vector ARMA processes
• Estimation of multivariate signal by output autocovariance data in linear discrete-time systems
• On the indirect approaches for CARMA model identification
• Robust spectral factor approximation of discrete-time frequency domain power spectras
• Subspace-based parameter estimation of symmetric noncausal autoregressive signals from noisy measurements
• A parametric estimation method for dynamic factor models of large dimensions
• USING SUBSPACE METHODS FOR ESTIMATING ARMA MODELS FOR MULTIVARIATE TIME SERIES WITH CONDITIONALLY HETEROSKEDASTIC INNOVATIONS
• A Suboptimum Maximum Likelihood Approach to Parametric Signal Analysis
• Stable reduced-rank VAR identification
• Modeling continuous-time processes via input-to-state filters

Describes a project that uses

Uses Software

• LMI toolbox

This page was built for publication: Vector ARMA estimation: A reliable subspace approach

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2734358)