Abstract
This paper suggests a new approach for fault detection using Genetic Algorithms (GAs). GAs are used to find the principal curve that summarize the data. The principal curve is a generation of linear Principal Component Analysis (PCA). Introduced by Hastie as a parametric curve, the original definition is based on the self-consistency property. The Hastie’s theory encloses weaknesses in case of complex data structures or data with intersections. The existing principal curves methods employ the first component of the data as an initial estimation of principal curve that passes satisfactorily through the middle of data. However the needing of an initial line is the major inconvenient of this approach. In this work, we extend this problem in two ways. First, we introduce a new method based on GAs to find the principal curve. Second, potential application of principal curves in fault detection is proposed. An example is presented to prove the efficiency of the proposed algorithm to fault detection of nonlinear process.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Nabli L (2010) Contribution à la conduite des systèmes de production par l’utilisation des techniques de l’intelligence artificielle. Habilitation universitaire, Université de Monastir, ENIM
Hastie T, Stuetzel W (1989) Principal curves. J Am Stat Assoc 84:502–512
Delicado P, Huerta M (2003) Principal curves of oriented points: theoretical and computational improvements. Comput Stat 18:293–315
Kègl B, Krzyzak A (2002) Piecewise linear skeletonization using principal curves. IEEE Trans Pattern Anal Mach Intell 24:59–74
Tibshirani R (1992) Principal curves revisited. Stat Comput 2:183–190
Kègl B, Krzyzak A, Linder T (2000) Learning and design of principal curves. IEEE Trans Pattern Anal Mach Intell 22:281–297
Banfield JD, Raftery AE (1992) Ice floe identification in satellite images using mathematical morphology and clustering about principal curves. J Am Stat Assoc 87:7–16
Chang K, Ghosh J (2001) A unified model for probabilistic principal surfaces. IEEE Trans Pattern Anal Mach Intell 23(1)
Einbeck J, Tutz G, Evers L (2005) Local principal curves. Stat Comput 15:301–313
Verbeek JJ, Vlassis N, Krose B (2002) A k-segments algorithm for finding principal curves. Pattern Recogn Lett 23:1009–1017
Goldberg D, Korb B, Deb K (1989) Messy genetic algorithms: motivation, analysis, and first results. J Complex Syst 3:493–530
Bies R, Muldoon F, Pollock G, Manuck S, Smith M (2006) A genetic algorithm-based hybrid machine learning approach to model selection. J Pharmacokinet Pharmacodyn 33:196–221
Cha SH, Tappert C (2009) A genetic algorithm for constructing compact binary decision trees. J Pattern Recogn Res 4:1–13
Akbari Z (2011) A multilevel evolutionary algorithm for optimizing numerical functions. Int J Ind Eng Comput 2:419–430
Zhang J, Chung H, Lo W (2007) Clustering-based adaptive crossover and mutation probabilities for genetic algorithms. IEEE Trans Evol Comput 11:326–335
Najeh T, Nabli L (2012) Development of a structuring residuals method for diagnostic by PCA-based genetic algorithms. In: The proceeding of 2nd international conference on communications, computing and control applications (CCCA), pp. 1–6, Marseilles, 6–8 Dec 2012
Nabli L, Ouni K, Haykel HS (2008) Approche Multi agents pour la surveillance indirecte d’un système de production par l’analyse en composantes principales. la Conférence Internationale Francophone d’Automatique, CIFA, Roumanie
Nabli L, Toguyéni AKA, Craye E (2000) Méthode de surveillance indirecte d’un système de production par la logique floue. CIFA, Lille
Kresta JV, MacGregor JF, Marlin TE (1991) Multivariate statistical monitoring of process operating performance. Can J Chem Eng 69:34–47
Nabli L, Ouni K (2008) The supervision indirect of a system of production by the principal component analysis and average dynamic of the metrics. Int Rev Autom Control (IREACO) 1(4):560–567
Delicado P (2001) Another look at principal curves and surface. J Multi-variate Anal 7:84–116
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer International Publishing Switzerland
About this paper
Cite this paper
Najeh, T., Telmoudi, A.J., Nabli, L. (2016). Nonlinear Process Monitoring Using Genetic Algorithms. In: Kumar, U., Ahmadi, A., Verma, A., Varde, P. (eds) Current Trends in Reliability, Availability, Maintainability and Safety. Lecture Notes in Mechanical Engineering. Springer, Cham. https://doi.org/10.1007/978-3-319-23597-4_9
Download citation
DOI: https://doi.org/10.1007/978-3-319-23597-4_9
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-23596-7
Online ISBN: 978-3-319-23597-4
eBook Packages: EngineeringEngineering (R0)