Application concept of artificial neural networks for turbo-machinery design

• Autorzy: Azzam M. , Haag J.-H. , Jeschke P.
• Języki publikacji: EN

Abstrakty

EN

This paper presents the results of an extensive investigation evaluating and improving the development of artificial neural network (ANN) models for turbomachinery design purposes. A set of 1100 differing axial compressor geometries based on 5 single-stage compressor rigs was prepared. Computations with the mean line analysis tool AXIAL™ took place to determine the according compressor maps defined by 15 operating points each. The challenge of ANN model development in terms of dimensionality reduction (feature selection), data normalization, defining the networks necessary plasticity, and network training is discussed using the example of three different models. As a result, the first model is able to predict the total pressure loss of the rotor blade row with a mean magnitude of the relative error (MMRE) of 3.6%. The second model predicts the total pressure ratio with an average accuracy of 0.8%. The third and last model was trained to predict basic geometrical parameters by presenting the load level and the performance data as an input. The achieved MMRE varied between 2.4% and 5.6% in respect of the particular output variable. The results show that ANNs are applicable to develop efficient models for turbomachinery design and analysis purposes, respectively.

Słowa kluczowe

EN

artificial neural network; turbomachinery; axial compressor

PL

sztuczna sieć neuronowa; turbomaszyna; sprężarka osiowa

• Wydawca: Instytut Podstawowych Problemów Techniki PAN
• Czasopismo: Computer Assisted Mechanics and Engineering Sciences
• Rocznik: 2009
• Tom: Vol. 16, No. 2
• Strony: 143--160
• Opis fizyczny: Bibliogr. 41 poz., tab., wykr.

Twórcy

autor

Azzam M.

autor

Haag J.-H.

autor

Jeschke P.

• lnstitute of Flight Systems and Automatic Control, Technische Universitdt Darmstadt, 64287 Darmstadt, Germany

Bibliografia

• [1] R. Ahmed. Auslegung vielstufiger Axialverdichter mit parallelen Evolutionsstrategien und neuronalen Netzen. PhD thesis, VDI Yerlag, Fortschritt-Berichte VDI Reihe 7, Stromungstechnik, NR.469, 2005.
• [2] A. Barron. Approximation and estimation bounds for artificial neural networks. In Workshop on Computational Learning Theory (COLT), 1991.
• [3] R. Bellman. Adaptive Control Processes: A Guided Tour. Princeton University Press, 1961.
• [4] E. Benini and A. Toffolo. Axial-flow compressor model based on a cascade stacking techniąue and neural networks. In Proceedings of the ASME Turbo Expo, June 2002.
• [5] C. M. Bishop. Neural Networks for Pattern Recognition. Oxford University Press, 1995.
• [6] C. M. Bishop. Pattern Recognition and Machine Learning. Springer Science+Business Media, LLC, 2006.
• [7] R. Callan. The Essence of Neural Networks. Prentice Hall, 1999.
• [8] M. V. Casey. Computational Methods for Preliminary Design and Geometry Definition in Turbomachinery. AGARD, LS-195, 1994.
• [9] M. Cetin, A. S. Ucer, Ch. Hirsch Ch., G. K. Serovy. Application of modified loss and deviation correlations to transonic axial compressors. AGARD Advisory Report No. 745, 1987.
• [10] R. V. Chima. Calculation of Tip Clearance Effects in a Transonic Compressor Rotor. NASA-TM-107216, 1996.
• [11] O. Dubitsky, A. Wiedermann, T. Nakano, J. Perera. The reduced order through-flow modeling of axial turbomachinery. In IGTC Tokyo TS-052, 2003.
• [12] R. Dunker. Test case eco-4-single transonic compressor stage. AGARD Adinsory Report, No. 275, 1990.
• [13] Zeli et al. SNNS: User Manual, Version A.2, 1998.
• [14] L. Fahrmeir, R. Kuenstler, L Pigeot, and G. Tutz. Statistik. Springer, Berlin Heidelberg New York, 6. edition, 2007.
• [15] R. B. Girider, D. Harris. Test case eco-3 single subsonic compressor stage. AGARD Advisory Report No. 275, 1990.
• [16] N. H.-R. Goerke. Reprasentation und Generierung von Robotertrajektorien mit neuronalen Netzen durch Einbau von Yorwissen. PhD thesis, Mathematisch-Naturwissenschaftliche Fakultat der Rheinischen Friedrich-Wilhelms-Uniyersitat Bonn, 1997.
• [17] I. Guyon, A. Elisseeff. An introduction to variable and feature selection. Journal of Machine Learning Research, 3:1137-1182, 2003.
• [18] J. Hartung, B. Elpelt. Multwariate Statistik. Oldenbourg Yerlag, 1992.
• [19] Ch. Hirsch, J. D. Denton. Through Flow Calculations in Axial Turbomachines. 1981.
• [20] Concepts ETI Inc. Axial User Gmde, Software Version 7.10.2. Concepts ETI Inc., 2000
• [21] META Group Inc. Data mining tools: Metaspectrum evaluation, September 2004.
• [22] N. Intrator, S. Edelman. Learning low-dimensional representations via the usage of multi-class labels. Comput. Neural Syst., 8:259-281, 1997.
• [23] M. Jansen. Globale Moddlbildung und garantiert stabile Regelung von Robotem mit Strukturierten Neuronalen Netzen. PhD thesis, VDI Yerlag, Fortschritt-Berichte VDI, 1995.
• [24] D. Japikse, N.C. Baines. Introduction to Turbomachinery. Concepts ETI, Inc. and Oxford University Press, Ist reprint edition, 1997.
• [25] R. Japikse, D. Japikse, R. Pelton. An evaluation of neural networks for meanline model deyelopment. In Proceedings of the ISROMAC, number 159, Honolulu, March 2004.
• [26] C. C. Koch, L. H. Smith. Loss sources and magnitudes in axial-flow compressors. Trans, of ASME: Journal of Engineering for Power, 98(3):411-424, July 1976.
• [27] S. Lieblein. Loss and stall analysis of compressor cascades. In Trans, of ASME: Journal of Basic Engineering, volume 81, pages 387-400, September 1959.
• [28] S. Lieblein. Incidence and deviation-arigle correlations for compressor cascades. In Trans, of ASME: Journal of Basic Engineering, volume 82, pages 575-587, 1960.
• [29] S. Lieblein, W.H. Roudebush. Theoretical loss relations for low-speed two-dimensional-cascade and isolated airfoil sections. Tn 3662, NAGA, 1956.
• [30] T. Masters. Practical Neural Network Recipes in C++. Academic Press, 1993.
• [31] S. Pazzi, F. Mertelli, V. Michelassi, M. Giachi. The use of artificial neural networks for performance prediction of return channels for industrial centrifugal compressors. In Proceedings of the ASME Turbo Expo, 2002.
• [32] S. Pierret, R.A. Van den Braembussche. Turbomachinery blade design using a navier-stokes solver and artificial neural network. ASME Trans., Journal of Turbomachinery, 121(2):326-332, 1999.
• [33] P. Rini, A. Alsalihi, R. A. Van den Braembussche. Evaluation of a design method for radial impellers based on artificial neural network and genetic algorithm. In Proceedings of the 5th ISAIF Conference, pp.535-543, Gdańsk, September 4-7 2001.
• [34] R. Rojas. Theorie der Neuronalen Netze. Springer-Verlag, 1993.
• [35] F. Rosenblatt. The perceptron: A probabilistic model for information storage and organization in the brain. Psychological Review, 65:386-408, 1958.
• [36] S. Schach, Th. Schafer. Regressions- und Yarianzanalyse. Springer-Yerlag, 1978.
• [37] H. Schneeweiss. Ekonometrie. MIT Press, 1986.
• [38] K. L. Suder. Blockage Development in a Transonic Axial Compressor Rotor. NASA-TM-113115, 1997.
• [39] M. Uelschen. Entwurf und Optimierung der zweidimensionalen Gitterstromung axialer Turbomschinenbeschaufelungen mit Neuronalen Netzen und Genetischen Algorithmen. PhD thesis, Kassel University Press, 2000.
• [40] A. Żeli. Simulation Neuronaler Netze. Addison-Wesley, 1994.
• [41] A. Żeli, I. Fischer, F. Hennecke, C. Bannes. Java NNS: User Manual, Version 1.1, 2001.