A PMF-based subspace method for continuous-time model identification. Application to a multivariable winding process

The identification and estimation of rainfall-flow models is one of the most challenging problems in hydrology. This paper presents the results of direct continuous-time identification and estimation of a rainfall-flow model for the Canning, an ephemeral river in Western Australia, based on daily sampled data. It compares simplified and full implementations of the optimal instrumental variable algorithm used in the application and discusses the advantages of this direct method when compared with the alternative indirect approach to the problem based on discrete-time model estimation.

show abstract

“…;Johansson, et al, 1999;Bastogne et al, 2001;Wang and Gawthrop, 2001;Garnier et al, 2003;Garnier and Young, 2004;Moussaoui et al, 2005).…”

Section: Optimal IV Methods For Ct Modelsmentioning

confidence: 99%

Identification and Estimation of Continuous-Time Rainfall-Flow Models

Young

Garnier

2006

IFAC Proceedings Volumes

View full text Add to dashboard Cite

show abstract

“…This can be done by augmenting the parameter vector in order to insert the extra parameters due to the initial conditions, or by starting the parameter estimation at t ¼ t 0 ; where t 0 is the time from which the function oðxð0Þ; i; tÞ can be neglected, as this function tends to zero when t increases. The time t 0 can be chosen as the time-response of the Poisson filter (Bastogne et al, 2001). The time-differential Eq.…”

Section: Article In Pressmentioning

confidence: 99%

“…With this toolbox, several works focused on estimating the performances of several methods with simulation system (Garnier, Mensler, & Richard, 2003) or laboratory-scaled systems (Mensler, Garnier, & Huselstein, 2000;Bastogne, Garnier, & Sibille, 2001) and, recently, a promising tentative of model identification via CT method of an industrial distillation column has been carried out (Garnier, 2002 Some CT methods provide very good performances and present the advantage that the estimated model is obtained directly in the CT domain without any model conversion as it would be the case with discrete-time identification methods. This is the other purpose of the paper: to show that the CT identification methods are now ready to be used successfully with actual industrial systems.…”

Section: Introductionmentioning

confidence: 99%

Identification of a toroidal continuously variable transmission using continuous-time system identification methods

Mensler

Joe

Kawabe

2006

Control Engineering Practice

View full text Add to dashboard Cite

“…It is an extension of the methods presented in the literature for rational (thus non fractional) systems [11][12][13] to the fractional case. Other subspace techniques for identifying continuous-time rational models can be found in [14][15][16][17]. So, the proposed method inherits the advantages of subspace methods which stem from the reliability of numerical algorithms using the QR and the singular value decompositions [18].…”

Section: Introductionmentioning

confidence: 99%