METAMODELING TECHNIQUES APPLIED TO AIRCRAFT NOISE PREDICTION

METAMODELING TECHNIQUES APPLIED TO AIRCRAFT NOISE PREDICTION

Aflalo, B. S.; Ferrari, D. B. T. P. A.

Full Article:

Given the increasing impact of aircraft noise for certification and commercialization purposes, a great effort has been made to decrease the acoustic emissions of newer aircraft. A great number of airports around the world limit noise emissions during the landing and takeoff phases, which contributes to increase the importance of controlling acoustic emissions of high-lift devices. Typically, simulations on aeroacustics are extremely expensive, because they involve fluid dynamics computa-tions in the transient regime. Not rarely, a single run of the computer code may require dozens of thousands of CPUs. Thus, direct optimization in aeroacoustics is often unfeasible for most of practical applications. A popular approach that has been currently adopted to overcome such difficulties involves creating a cheap-to-compute, yet accurate, surrogate (or metamodel) for the computation-intensive function on which optimization is to be performed. This work aims at using and analyzing the performance of various metamodeling techniques in slat noise simula-tions, carried out using a 2D discrete vortices method. Three classes of metamodels, were tested: (i) first and second order polynomial regressions, (ii) artificial neural networks, and (iii) Gaussian stochastic processes (GaSP). For the neural networks model, various models were created in order to verify which combinations of numbers of neurons and layers yielded better results. For the GaSP models, different covariance functions and base functions were tested. All metamodels were constructed based on data generated according to a Latin hypercube design with 28 points in a four-dimensional input space consisting in angle of attack, slat deflection angle and the two dimensions of slat position. The performance of each model was assessed with respect to the root mean square cross-validation prediction error. Results for the performed simulations indicated that the Gaussian stochastic process with the power expo-nential covariance function and a first order base function produced the best metamodel.

Given the increasing impact of aircraft noise for certification and commercialization purposes, a great effort has been made to decrease the acoustic emissions of newer aircraft. A great number of airports around the world limit noise emissions during the landing and takeoff phases, which contributes to increase the importance of controlling acoustic emissions of high-lift devices. Typically, simulations on aeroacustics are extremely expensive, because they involve fluid dynamics computa-tions in the transient regime. Not rarely, a single run of the computer code may require dozens of thousands of CPUs. Thus, direct optimization in aeroacoustics is often unfeasible for most of practical applications. A popular approach that has been currently adopted to overcome such difficulties involves creating a cheap-to-compute, yet accurate, surrogate (or metamodel) for the computation-intensive function on which optimization is to be performed. This work aims at using and analyzing the performance of various metamodeling techniques in slat noise simula-tions, carried out using a 2D discrete vortices method. Three classes of metamodels, were tested: (i) first and second order polynomial regressions, (ii) artificial neural networks, and (iii) Gaussian stochastic processes (GaSP). For the neural networks model, various models were created in order to verify which combinations of numbers of neurons and layers yielded better results. For the GaSP models, different covariance functions and base functions were tested. All metamodels were constructed based on data generated according to a Latin hypercube design with 28 points in a four-dimensional input space consisting in angle of attack, slat deflection angle and the two dimensions of slat position. The performance of each model was assessed with respect to the root mean square cross-validation prediction error. Results for the performed simulations indicated that the Gaussian stochastic process with the power expo-nential covariance function and a first order base function produced the best metamodel.

Palavras-chave:

DOI: 10.5151/meceng-wccm2012-20194

Referências bibliográficas

• [1] ICAO - International Civil Aviation Organization. Disponível em: http://www.icao.int/environmental-protection/Pages/default.aspx. Acessado em: 25/01/2012.
• [2] Choudhari, M. M., Khorrami, M. R., Effect of Three-Dimensional Shear-Layer Structures on Slat Cove Unsteadiness, AIAA Journal Vol 45, No.9, September 2007.
• [3] Aflalo, B. S., Simoes, L. G. C., Silva, R. G., Medeiros, M. A. F., “Comparative Analysis of Turbulence Models for Slat Noise Source Calculations Employing Unstructured Mesh-es”, 16th AIAA/CEAS Aeroacoustics Conference, Stocolmo, Suécia. AIAA-2010-3838.
• [4] Guo, Y., "A Discrete Vortex Model for Slat Noise Prediction", 7th AIAA/CEAS Aeroa-coustics Conference, Maastricht, Netherlands. AIAA-2001-2157.
• [5] Ffowcs Williams, J. E., and Hawkings, D. L., "Sound Generated by Turbulence and Sur-faces in Arbitrary Motion", Philosophical Transactions of the Royal Society, Vol. A264, 1969, pp. 321-342.
• [6] Rackl, R. G., Miller, G., Guo, Y., Yamamoto, K., "Airframe Noise Studies - Review and Future Direction".NASA/CR-2005-213767.
• [7] Ferrari, D. Andamp; Head, T., Regression in R Part I :Simple Linear Regression, UCLA De-partment of Statistics - Statistical Consulting Center, 2010.
• [8] Least Squares. Disponível em: http://en.wikipedia.org/wiki/Least_squares. Acessado em: 1º de dezembro de2011.
• [9] Stergiou, C.; Siganos, D., Neural Networks. Disponível em: http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.html#. What is a Neural Network. Acessado em: 07 de setembro de 2010.
• [10] Ferrari, D., B., T., P., A., "Multi-fidelity Data Fusion in Statistical Metamodeling of Aero-dynamic Functions". PhD thesis in Statistics, University of California, Los Angeles, 2010.
• [11] Sacks, J., Schiller, S. B. and Welch, W. J. (1989), “Designs for computer experiments”, Technometrics 31, 41–47.
• [12] Cook''s Distance, Wikipedia, the Free Encyclopedia. Acessado em 2 de dezembro de 2011, Disponível em http://en.wikipedia.org/wiki/Cook''s_distance.
• [13] The AMORE package: A MORE flexible neural network package. Disponível em: http://rwiki.sciviews.org/doku.php?id=packages:cran:amore. Acessado em: 12/12/2011.
• [14] Roustant, O., Ginsbourger, D., Deville, Y., "DiceKriging, DiceOptim: Two R packages for the analysis of computer experiments by kriging-based metamodeling and optimiza-tion". HAL-00495766, versão 1 - 28 Jun 2010.

Como citar:

Aflalo, B. S.; Ferrari, D. B. T. P. A.; "METAMODELING TECHNIQUES APPLIED TO AIRCRAFT NOISE PREDICTION", p-5018-5033. In: In Proceedings of the 10th World Congress on Computational Mechanics [= Blucher Mechanical Engineering Proceedings, v. 1, n. 1]. São Paulo: Blucher, 2014.
ISSN 23580828, DOI 10.5151/meceng-wccm2012-20194