Int. J. of Adaptive and Innovative Systems   »   2010 Vol.1, No.3/4

 

 

Title: Multiple sensor fault detection and isolation of an air quality monitoring network using RBF-NLPCA model

 

Author: Mohamed-Faouzi Harkat, Yvon Tharrault, Gilles Mourot, Jose Ragot

 

Addresses:
University Badji Mokhtar – Annaba, Faculty of Engineering Sciences, Department of Electronic, P.O. Box 12, Annaba, 23000, Algeria.
Centre de Recherche en Automatique de Nancy (CRAN), Nancy University, CNRS UMR 7039, 2 avenue de la foret de Haye, F-54516 Vandoeuvre-Les-Nancy, France.
Centre de Recherche en Automatique de Nancy (CRAN), Nancy University, CNRS UMR 7039, 2 avenue de la foret de Haye, F-54516 Vandoeuvre-Les-Nancy, France.
Centre de Recherche en Automatique de Nancy (CRAN), Nancy University, CNRS UMR 7039, 2 avenue de la foret de Haye, F-54516 Vandoeuvre-Les-Nancy, France

 

Abstract: This paper presents a data-driven method based on non-linear principal component analysis to detect and isolate multiple sensor faults. The RBF-NLPCA model is obtained by combining a principal curve algorithm and two three-layer radial basis function (RBF) networks. The reconstruction approach for multiple sensors is proposed in the non-linear case and successfully applied for multiple sensor fault detection and isolation of an air quality monitoring network. The proposed approach reduces considerably the number of reconstruction combinations and allows to determine replacement values for the faulty sensors.

 

Keywords: multiple fault detection; fault isolation; nonlinear PCA; principal component analysis; radial basis functions; RBF; reconstruction; air quality monitoring networks; air pollution; multiple sensors; replacement values; faulty sensors; neural networks.

 

DOI: 10.1504/IJAIS.2010.034804

 

Int. J. of Adaptive and Innovative Systems, 2010 Vol.1, No.3/4, pp.267 - 284

 

Available online: 23 Aug 2010

 

 

Editors Full text accessAccess for SubscribersPurchase this articleComment on this article