Global total least squares modeling of multivariable time series

1995/01/01 by B. Roorda, Berend Roorda, C. Heij +1 · 2 citations
Physics and Astronomy · Engineering · Decision Sciences · #Model Reduction and Neural Networks #Control Systems and Identification #Probabilistic and Robust Engineering Design

paper · doi:10.1109/9.362900

Abstract

Presents a novel approach for the modeling of multivariable time series. The model class consists of linear systems, i.e., the solution sets of linear difference equations. Restricting the model order, the aim is to determine a model with minimal l/sub 2/-distance from the observed time series. Necessary conditions for optimality are described in terms of state-space representations. These conditions motivate a relatively simple iterative algorithm for the nonlinear problem of identifying optimal models. Attractive aspects of the proposed method are that the model error is measured globally, it can be applied for multi-input, multi-output systems, and no prior distinction between inputs and outputs is required. The authors give an illustration by means of some numerical simulations.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Citations

• Principal component analysis in linear systems: Controllability, observability, and model reduction

Cited by

• Control-Oriented System Identification: Classical, Learning, and Physics-Informed Approaches
• Robustness of Maximum Correntropy Estimation Against Large Outliers

Related

• Windowed least-squares model reduction for dynamical systems
• On model reduction by least squares moment matching
• Model reduction by least squares moment matching for linear and nonlinear systems
• Derivation of an EM algorithm for constrained and unconstrained multivariate autoregressive state-space (MARSS) models
• Least Squares Model Reduction: A Two-Stage System-Theoretic Interpretation