Identification of linear parameter-varying systems: A reweighted ℓ2,1-norm regularization approach

•Nonlinear least-squares identification of linear parameter-varying systems is addressed.•Reweighted ℓ2,1-norm regularization is applied to trade-off model accuracy against model simplicity.•The resulting nonsmooth optimization is solved using a sequential convex programming approach.•The method is...

Full description

Saved in:
Bibliographic Details
Published in:Mechanical systems and signal processing 2018-02, Vol.100, p.729-742
Main Authors: Turk, D., Gillis, J., Pipeleers, G., Swevers, J.
Format: Article
Language:English
Subjects:
Basis selection
Convexity
Iterative methods
Least squares
Linear parameter-varying (LPV) systems
Mathematical models
Model accuracy
Optimization
Regularization
Regularization methods
State-space models
System identification
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•Nonlinear least-squares identification of linear parameter-varying systems is addressed.•Reweighted ℓ2,1-norm regularization is applied to trade-off model accuracy against model simplicity.•The resulting nonsmooth optimization is solved using a sequential convex programming approach.•The method is successfully validated using numerical and experimental data. This paper presents a regularized nonlinear least-squares identification approach for linear parameter-varying (LPV) systems. The objective of the method is, on the one hand, to obtain an LPV model of which the response fits the system measurements as accurately as possible and, on the other hand, to favor models with an as simple as possible dependency on the scheduling parameter. This is accomplished by introducing ℓ2,1-norm regularization into the nonlinear least-squares problem. The resulting nonsmooth optimization problem is reformulated into a nonlinear second-order cone program and solved using a sequential convex programming approach. Through an iterative reweighting of the regularization, the parameters that do not substantially contribute to the system response are penalized heavily, while the significant parameters remain unaffected or are penalized only slightly. Numerical and experimental validations of the proposed method show a substantial model simplification in comparison with the nonregularized solution, without significantly sacrificing model accuracy.
ISSN:0888-3270
1096-1216
DOI:10.1016/j.ymssp.2017.07.054