CCE Faculty Articles
Title
Identification of LPV State Space Systems by a Separable Least Squares Approach
Event Date/Location
Florence, Italy
Document Type
Article
Date
12-2013
Publication Title
IEEE Conference on Decision and Control
ISSN or ISBN
0191-2216
First Page
4104
Last Page
4109
Description
In this article, an algorithm to identify LPV State Space models is proposed. The LPV State Space system is in the companion reachable canonical form. Both the state matrix and the output vector coefficients are linear combinations of a set of nonlinear basis functions dependent on the scheduling signal. This model structure, although simple, can describe accurately the behaviour of many nonlinear systems by an adequate choice of the scheduling signal. The identification algorithm minimises a quadratic criterion of the output error. Since this error is a linear function of the output vector parameters, a separable nonlinear least squares approach is used to minimise the criterion function by a gradient method. The derivatives required by the algorithm are the states of LPV systems that need to be simulated at every iteration. The effectiveness of the algorithm is assessed by two simulated examples.
DOI
10.1109/CDC.2013.6760518
NSUWorks Citation
Ramos, Jose A.; Lopes dos Santos, Paulo; Azevedo Perdicoulis, Teresa Paula; Martins de Carvalho, Jorge L.; and Rivera, Daniel E., "Identification of LPV State Space Systems by a Separable Least Squares Approach" (2013). CCE Faculty Articles. 410.
https://nsuworks.nova.edu/gscis_facarticles/410
