Call for Paper - March 2023 Edition
IJCA solicits original research papers for the March 2023 Edition. Last date of manuscript submission is February 20, 2023. Read More

The Effect of Nonlinearity Measure on Model Order Selection in Identification of Chemical Processes

Print
PDF
International Journal of Computer Applications
© 2011 by IJCA Journal
Number 1 - Article 1
Year of Publication: 2011
Authors:
Mohammad Rezaee Yazdi
Hooshang Jazayerirad
10.5120/3929-5134

Mohammad Rezaee Yazdi and Hooshang Jazayerirad. Article:The Effect of Nonlinearity Measure on Model Order Selection in Identification of Chemical Processes. International Journal of Computer Applications 32(8):1-5, October 2011. Full text available. BibTeX

@article{key:article,
	author = {Mohammad Rezaee Yazdi and Hooshang Jazayerirad},
	title = {Article:The Effect of Nonlinearity Measure on Model Order Selection in Identification of Chemical Processes},
	journal = {International Journal of Computer Applications},
	year = {2011},
	volume = {32},
	number = {8},
	pages = {1-5},
	month = {October},
	note = {Full text available}
}

Abstract

Most processes in chemical industry reveal nonlinear behavior. A key requirement for many advanced process control is the availability of an accurate dynamic model, with lowest order, for process. This paper handles the problem of model order selection in non-linearity chemical processes identification procedure. In this respect a ‘nonlinearity test’ method and four model order selection criteria known as the Aikaike Information Criterion (AIC), the Minimum Description Length (MDL), the Exponentially Embedded Family (EEF) and Unmodeled Output Variation (UOV) are considered. The abilities of these criteria in determining the order of the model subjected to different levels of nonlinearity are compared. For this purpose, two chemical processes: a two-tank system and a continuous stirred tank reactor (CSTR) with different levels of nonlinearity are employed. It has been shown that in a system with high level of nonlinearity, the UOV criterion is able to select the lowest order model compared to the other criteria.

Reference

  • Guay. M. 1996 Measurement of Nonlinearity in Chemical Process Control. Doctoral Thesis. UMI Order No. 1012., Department of Chemical Engineering, Queen’s University at Kingston.
  • Lind. I. 2006 Regressor and Structure Selection Uses of ANOVA in System Identification. Doctoral Thesis. UMI Order No.895., Department of Electrical Engineering, Linkoping University.
  • Ridder, F.D. Pintelon, R. Schoukens, J., Gillikin, D.P., “Modified AIC and MDL Model Selection Criteria for Short Data Record”, (IEEE) TIM, 2005.
  • Haber, R. 1985. Nonlinearity Tests for Dynamic Processes. Proc.IFAC Symp.
  • Tong, H. 1990 Nonlinear Time Series, a Dynamical System Approach. Oxford UK Clarendon Press.
  • Stark. J.A, “Parameter-based Hypothesis Tests for Model Selection”, Signal Processing (EURASIP), 1995.
  • Bindlish. R, “Model order selection for process identification applied to an industrial ethylene furnace”, Jprocont, 2003.
  • Ding. Q, Kay. S, “Inconsistency of the MDL: On the Performance of Model Order Selection Criteria with Increasing Signal-to-Noise Ratio”, Signal Processing (IEEE), 2011.
  • MALAB, “System Identification Toolbox Documentation’, Mathworks Inc Natick, MA., 2010.
  • Luyben, M.L. Luyben, W.L. 1997 Essentials of Process Control. McGraw-Hill Singapore.