PCA-based approximation of a class of distributed parameter systems: classical vs. neural network approach

• Autorzy: Bartecki K.
• Języki publikacji: EN

Abstrakty

EN

In this article, an approximation of the spatiotemporal response of a distributed parameter system (DPS) with the use of the principal component analysis (PCA) is considered. Based on a data obtained by the numerical solution of a set of partial differential equations, a PCA-based approximation procedure is performed. It consists in the projection of the original data into the subspace spanned by the eigenvectors of the data covariance matrix, corresponding to its highest eigenvalues. The presented approach is carried out using both the classical PCA method as well as two different neural network structures: two-layer feed-forward network with supervised learning (FF-PCA) and single-layer network with unsupervised, generalized Hebbian learning rule (GHA-PCA). In each case considered, the effect of the approximation model structure represented by the number of eigenvectors (or, in the neural case, units in the network projection layer) on the mean square approximation error of the spatiotemporal response and on the data compression ratio is analysed. As shown in the paper, the best approximation quality is obtained for the classical PCA method as well as for the FF-PCA neural approach. On the other hand, an adaptive learning method for the GHA-PCA network allows to use it in e.g. an on-line identification scheme.

Słowa kluczowe

EN

distributed parameter system; principal component analysis; artificial neural network; supervised learning; unsupervised learning

• Wydawca: Polska Akademia Nauk, Wydział IV Nauk Technicznych
• Czasopismo: Bulletin of the Polish Academy of Sciences. Technical Sciences
• Rocznik: 2012
• Tom: Vol. 60, nr 3
• Strony: 651--660
• Opis fizyczny: Bibliogr. 35 poz., rys., tab.

Twórcy

autor

Bartecki K.

• Institute of Control and Computer Engineering, Opole University of Technology, 31 Sosnkowskiego St., 45-272 Opole, Poland

Bibliografia

• [1] K. Bartecki, “Optimization of the night setback process for a heated building using neural algorithms”, Proc. 5th IEEEInt. Conf. on Neural Networks and Soft Computing 1, 675-680 (2000).
• [2] K. Bartecki, “Comparison of frequency responses of paralleland counter-flow type of heat exchanger”, Proc. 13th IEEEIFAC Int. Conf. on Methods and Models in Automation andRobotics 1, CD-ROM (2007).
• [3] K. Bartecki, “Frequency- and time-domain analysis of a simple pipeline system”, Proc. 14th IEEE IFAC Int. Conf. on Methodsand Models in Automation and Robotics 1, CD-ROM (2009).
• [4] K. Bartecki and R. Rojek, “Performance optimization of heat exchanging centre using neural algorithms”, Proc. 4th IEEEInt. Conf. on Neural Networks and Their Applications 1, 643- 648 (1999).
• [5] K. Bartecki and R. Rojek, “Instantaneous linearization of neural network model in adaptive control of heat exchange process”, Proc. 11th IEEE Int. Conf. on Methods and Modelsin Automation and Robotics 1, 967-972 (2005).
• [6] J.C. Friedly, Dynamic Behaviour of Processes, Prentice Hall, New York, 1972.
• [7] H.-X. Li and C. Qi, “Modeling of distributed parameter systems for applications - a synthesized review from time-space separation”, J. Process Control 20, 891-901 (2010).
• [8] B. Cichy, P. Augusta, E. Rogers, K. Gałkowski, and Z. Hur´ak, “Robust control of distributed parameter mechanical systems using a multidimensional systems approach”, Bull. Pol. Ac.:Tech. 58 (1), 67-75 (2010).
• [9] K.Murawski and D. Lee, “Numerical methods of solving equations of hydrodynamics from perspectives of the code flash”, Bull. Pol. Ac.: Tech. 59 (1), 81-91 (2011).
• [10] R. Rojek, Modeling of Flow Networks, Opole University of Technology Press, Opole, 2002.
• [11] L.G. Bleris and M.V. Kothare, “Low-order empirical modeling of distributed parameter systems using temporal and spatial eigenfunctions”, Computers and Chemical Engineering 29 (4), 817-827 (2005).
• [12] K.A. Hoo and D. Zheng, “Low-order control-relevant models for a class of distributed parameters system”, ChemicalEngineering Science 56 (23), 6683-6710 (2001).
• [13] Qi Ch. and H.-X. Li, “A time-space separation-based Hammerstein modeling approach for nonlinear distributed parameter processes”, Computers and Chemical Engineering 33 (7), 1247-1260 (2009).
• [14] K. Bartecki, “Approximation of a class of distributed parameter systems using proper orthogonal decomposition”, Proc.16th IEEE Int. Conf. on Methods and Models in Automationand Robotics 1, CD-ROM (2011).
• [15] G. Berkooz, P. Holmes, and J.L. Lumley, “The proper orthogonal decomposition in the analysis of turbulent flows”, AnnualReview of Fluid Mechanics 25, 539-575 (1993).
• [16] A. Chatterjee. “An introduction to the proper orthogonal decomposition”, Current Science 78 (7), 808-817 (2000).
• [17] Y.C. Liang, H.P. Lee, S.P. Lim, W.Z. Lin, K.H. Lee, and C.G. Wu, “Proper ortogonal decomposition and its applications - part I: Theory”, J. Sound and Vibration 252 (3), 527-544 (2002).
• [18] H.M. Park and D.H. Cho, “The use of the Karhunen-Lo`eve decomposition for the modeling of distributed parameter systems”, Chemical Engineering Science 51 (1), 81-98 (1996).
• [19] C. Wolter, M.A. Trindade, and R. Sampaio, “Obtaining mode shapes through the Karhunen-Lo`eve expansion for distributedparameter linear systems”, Shock and Vibration 9 (4), 177-192 (2002).
• [20] G. Kerschen, F. Poncelet, and J.-C. Golinval, “Physical interpretation of independent component analysis in structural dynamics”, Mechanical Systems and Signal Processing 21 (4), 1561-1575 (2007).
• [21] Y.C. Liang, H.P. Lee, S.P. Lim, W.Z. Lin, K.H. Lee, and H. Sun, “Proper orthogonal decomposition and its applications - part II: Model reduction for MEMS dynamical analysis”, J.Sound and Vibration 256 (3), 515-532 (2002).
• [22] D. Zheng and K. A. Hoo, “Low-order model identifcation for implementable control solutions of distributed parameter systems”, Computers and Chemical Engineering 26, 1049-1076 (2002).
• [23] M. Li and P.D. Christofides, “Optimal control of diffusionconvection- reaction processes using reduced-order models”, Computers and Chemical Engineering 32, 2123-2135 (2007).
• [24] E. Aggelogiannaki and H. Sarimveis, “Nonlinear model predictive control for distributed parameter systems using data driven artificial neural network models”, Computers & Chemical Engineering 32 (6), 1225-1237 (2008).
• [25] K.I. Diamantaras and S.Y. Kung, Principal Component Neural Networks - Theory and Applications, John Wiley, New York, 1996.
• [26] E. Oja, “A simplified neuron model as a principal component analyzer”, J. Mathematical Biology 15 (3), 267-273 (1982).
• [27] E. Oja, “Neural networks, principal components, and subspaces”, Int. J. Neural Systems 1 (1), 61-68 (1989).
• [28] S. Osowski, Neural Networks for Information Processing, Warsaw University of Technology Press, Warsaw, 2000.
• [29] T.D. Sanger, “Optimal unsupervised learning in a single-layer linear feedforward neural network”, Neural Networks 2 (6), 459-473 (1989).
• [30] J. Żurada, Introduction to artificial neural systems, West Publishing Company, St. Paul, 1992.
• [31] K. Bartecki, “On some peculiarities of neural network approximation applied to the inverse kinematics problem”, Proc. Conf.on Control and Fault-Tolerant Systems 1, 317-322 (2010).
• [32] J. Hertz, A. Krogh, and R. Palmer, Introduction to the Theoryof Neural Computation, Addison Wesley, Redwood City, 1991.
• [33] R. Rojas, Neural Networks: a Systematic Introduction, Springer, Berlin, 1996.
• [34] M. Ziołko, Modeling of Wave Phenomena, AGH University of Science and Technology Press, Krakow, 2000.
• [35] R.M.M. Mattheij, S.W. Rienstra, and J.H.M. ten Thije Boonkkamp, Partial Differential Equations: Modeling,Analysis, Computation, Society for Industrial and Applied Mathematics (SIAM), Philadelphia, 2005.