3-D AERODİNAMİK OPTİMİZASYON İÇİN HİBRİT BİR YÖNTEMİN (ANN ve GA) YENİ KULLANIMI ÜZERİNE

Bu çalışmanın amacı, genetik algoritma ve yapay sinir ağını kullanarak 3 boyutlu kanat konfigürasyonları için daha verimli ve etkin bir hibrid aerodinamik optimizasyon metodu sunmaktır. Yapay Sinir Ağı (ANN) ileri ok açılı kanadın aerodinamik optimizasyonunda yeni bir yaklaşımla kullanılmıştır. Geliştirilen tekniğin Genetik Algoritma (GA) yöntemlerinden çok daha etkin ve güçlü olduğu görülmüştür. Örneğin, yeni hibrid teknik, sadece GA kullanan yöntemin 500 akış çözümüyle ulaştığı seviyeye, yaklaşık yarı zamanda (250 hesaplamada) ulaşabilmektedir. Sürükleme katsayısında sağlanan azalma, önerilen yöntemde % 33 daha hızlı sağlanmaktadır. Yapay sinir ağı algoritması, olası kötü üyeleri önlemek için düzenlenmiş olan özel bir elitizm yöntemiyle birlikte GA içine eklenmiştir.

ON NOVEL USAGE OF A HYBRID METHOD (ANN and GA) FOR FASTER 3-D AERODYNAMIC OPTIMIZATION

The purpose of this study is to offer a more efficient hybrid aerodynamic optimization method for 3-D wing configurations by using both genetic and artificial neural network. Artificial Neural Network (ANN) is used with a new approach in the aerodynamic optimization of a forward swept wing. The developed technique has been found much more robust than Genetic Algorithm (GA) only methods. For example, the new hybrid technique acquires the same fitness level as the one that GA only method can reach in 500 calculations, in about half time (about 250 calculations). The drag coefficient reduction is calculated %33 faster in the offered method. The neural network is embedded into the genetic algorithm along with augmented elitism to prevent possible bad members in the generations.

PDF

___

[1] Foster, N.F. and Dulikravich, G.S., Three Dimensional Aerodynamic Shape Optimization Using Genetic and Gradient Search Algorithms, Journal of Spacecraft and Rockets, Vol. 34, No. 1, 1997
[2] Hacıoğlu A. and Özkol İ., (2003), Transonic Airfoil Design and Optimization by Using Vibrational Genetic Algorithm, Aircraft Engineering and Aerospace Technology, Vol.75, No:4.
[3] Liu, L.L., (2005), "Intelligent Genetic Algorithm and Its Application to Aerodynamic Optimization of Airplanes" AIAA Journal, Vol. 43, No. 3.
[4] Hacıoğlu, A., (2005), "A novel usage of neural network in optimization and implementation to the internal flow systems", Aircraft Engineering and Aerospace Technology, Vol. 77, No. 5, pp. 369-375.
[5] Epstein, B., Peigin, S. (2004), "Robust Hybrid Approach to Multiobjective Constrained Optimization in Aerodynamics", AIAA Journal, Vol. 42, No. 8, pp. 1572-1579.
[6] Qazi, M., and Linshu, M. (2006), "Nearlyorthogonal sampling and neural network metamodel driven conceptual design of multistage space launch vehicle", Computer-Aided Design, Vol. 38, No. 6, pp. 595-607.
[7] Ludwig, O., Nunes, U., Araújo, R., Schnitman, L., Lepikson, H.A. (2009), "Applications of information theory, genetic algorithms, and neural models to predict oil flow, Communications in Nonlinear Science and Numerical Simulation, Vol. 14, No. 7, pp. 2870-2885.
[8] Kurtulus, D.F. (2009), "Ability to forecase unsteady aerodynamic forces of flapping airfoils by artificial neural network", Neural Computing & Applications, Vol. 18 Issue 4, pp. 359-368.
[9] Huang, J.T., Gao, Z.H., Zhao, K., Ba, J.Q. (2010), "Robust design of supercritical wing aerodynamic optimization considering fuselage interfering", Chinese Journal of Aeronautics, October 2010, 23 5, pp. 523-538.
[10] Sun, G., Sun, Y., Wang, S. (2015), "Artificial neural network based inverse design: Airfoils and wings", Aerospace Science and Technology, AprilMay 2015.
[11] Yilmaz, E., Kavsaoglu, M.S., Akay, H.U. and Akmandor, I.S., (2001), "Cell-Vertex Based Parallel and Adaptive Explicit 3D Flow Solution on Unstructured Grids", The International Journal of CFD, Vol. 14, pp.271-286
[12] Vatandaş, E., Özkol İ., and Hacıoğlu, A. (2007), "Vibrational Genetic Algorithm (VGA) and Dynamic Mesh in the Optimization of 3-D Wing Geometries" Journal of Inverse Problems in Science and Engineering, Vol. 15, No. 6, pp. 643-657.
[13] Vatandaş, E., (2007a), "Geometrical and Positional Optimization of the Forward Swept Lift Producing Surfaces in 3-D Flow Domains" Aircraft Engineering & Aerospace Technology, Vol.79, No.6, pp. 635-645.
[14] Vatandaş, E., Özkol İ., (2008) "Coupling Dynamic Mesh Technique and Heuristic Algorithms in 3 Dimensional Tapered Wing Design" International Journal of Numerical Methods in Engineering, Vol. 74 No. 12, pp. 1771 - 1794.