Some Study on the Identification of Multi-Model LPV Models with Two Scheduling Variables

Huang, Jiangyin; Ji, Guoli; Zhu, Yucai; Bosch, Paul

doi:10.3182/20120711-3-be-2027.00008

Cited by 2 publications

(5 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On the other hand, the "glocal" method using trapezoidal approach enables preserving the local properties of the model fairly well, as the basis functions have a region where only a single model is active and the influence of the other local models is limited (unlike in the polynomial approach). This explains the contradictory results reported in [27,35] using the polynomial interpolation method. However, it is important to highlight that the problem is not with the polynomial method, but with the "glocal" nature of the applied scheme.…”

Section: Assessment Of the Resultscontrasting

confidence: 43%

“…This property, called changing local model order, jeopardizes any local identification approach, which is based on interpolation of IO or state-space models. It also explains why the approach reported in [35] is applied successfully to other applications without this property. In the LPV-OBF identification approach, the flexible series expansion model structure is resilient to the changing local model order behavior, leading to promising results in both model simulation fitness and low local H 2 error ratio.…”

Section: Assessment Of the Resultsmentioning

confidence: 84%

“…a particular trajectory of p may not fully reveal the quality of the identified models. A similar observation for the distillation column has been obtained in [35,27]. To systematically verify whether the models can generalize the local dynamics of the system w.r.t.…”

Section: Assessment Of the Resultsmentioning

confidence: 99%

“…Selection of the model structure and the interpolation scheme as well as the number and location of the used operating points are application specific (e.g., see [10]). Some of the local approaches have been recently applied on distillation columns [35,27], on fermentation processes [36] and on a continuous tank reactor [28,36]. In the sequel, we intend to give an overview on all available methods for the LPV modeling of process systems in the local setting.…”

Section: Lpv Identification Using the Local Approachmentioning

confidence: 99%

“…The primary motivation for investigating the above mentioned local and global approaches for data-driven modeling of a highpurity distillation column originates from [35,27], where significant difficulties are reported for the local LPV identification approaches. The objective of this work is twofold: (i) provide a complete overview of the local LPV approaches given in [27] and (ii) investigate the open questions related to the performance of local and global LPV approaches.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

A review on data-driven linear parameter-varying modeling approaches: A high-purity distillation column case study

Bachnas

Tóth

Ludlage

et al. 2014

Journal of Process Control

108

View full text Add to dashboard Cite

a b s t r a c tModel-based control strategies are widely used for optimal operation of chemical processes to respond to the increasing performance demands in the chemical industry. Yet, obtaining accurate models to describe the inherently nonlinear, time-varying dynamics of chemical processes remains a challenge in most model-based control applications. This paper reviews data-driven, Linear Parameter-Varying (LPV) modeling approaches for process systems by exploring and comparing various identification methods on a high-purity distillation column case study. Several LPV identification methods that utilize input-output and series expansion model structures are explored. Two LPV identification perspectives are adopted: (i) the local approach, which corresponds to the interpolation of Linear Time-Invariant (LTI) models identified at different steady-state operating points of the system and (ii) the global approach, where a parametrized LPV model structure is identified directly using a global data set with varying operating points. For the local approach, various model interpolation schemes are studied under an Output Error (OE) noise setting, whereas in the global case, a polynomial parametrization based OE prediction error minimization approach, an Orthonormal Basis Functions (OBFs) based model estimator and a LeastSquare Support Vector Machine (LS-SVM) based non-parametric approach are investigated. Through extensive simulation studies, the aforementioned LPV identification approaches are analyzed in terms of the attainable model accuracy and local frequency response behavior of the obtained models. Recommendations are provided to achieve adequate choice between the methods for a particular process system at hand.

show abstract

Section: Assessment Of the Resultscontrasting

confidence: 43%

Section: Assessment Of the Resultsmentioning

confidence: 84%

Section: Assessment Of the Resultsmentioning

confidence: 99%