Model reduction and prediction of chaotic time series for model based control of crystallizers

J. Abonyi, B. Feil, B. G. Lakatos

Research output: Contribution to journalArticle

Abstract

The paper proposes a new algorithm for model-reduction of crystallizers that is based on fuzzy clustering of chaotic time series generated by continuous crystallization processes under some operation conditions. To determine the proper number of the state variables, the so-called embedding dimension, the clustering is applied in the reconstructed space defined by the lagged variables whose values are measurable. The correct embedding dimension is inferred from the one-step-ahead prediction performance of the local models of the clusters. The main advantage of the proposed solution is that three tasks are simultaneously solved during clustering: selection of the embedding dimension, estimation of the intrinsic dimension, and identification of a model that can be used for prediction of chaotic time series. The usefulness of the method is demonstrated by modelling two coupled continuous MSMPR crystallizers.

Original languageEnglish
Article numberE-4
Pages (from-to)1009-1014
Number of pages6
JournalVDI Berichte
Issue number1901 II
Publication statusPublished - Nov 7 2005

ASJC Scopus subject areas

  • Engineering(all)

Fingerprint Dive into the research topics of 'Model reduction and prediction of chaotic time series for model based control of crystallizers'. Together they form a unique fingerprint.

  • Cite this