An unsteady aerodynamic surrogate model based on the deep LSTM (long short-term memory) network is proposed for predicting unsteady aerodynamic coefficients. Deflection angles and deflection velocities of control surfaces are introduced to input values of the surrogate model to enhance the capability of identifying different motion states so that accumulative error can be controlled. Longitudinal stability is extremely important for flight safety while few studies have worked on unsteady aerodynamics of airfoils/wings with moving high-lift device (HLD) motion. Longitudinal static margin sequence of the HLD extending process is studied, and nonlinearity and hysteresis of coefficients in HLD motion are validated. The surrogate model is then applied in HLD motion control with the particle swarm optimization (PSO) method. Additionally, the results are then performed in three-dimensional aircraft HLD control. Validation computations show that longitudinal stability of optimized configuration is promoted with lift coefficient unchanged.
PehlivanogluYV. Efficient accelerators for PSO in an inverse design of multi-element airfoils. Aerospace Sci Tech2019; 91: 110–121.
2.
SoulatLPouanguéAFMoreauS. A high-order sensitivity method for multi-element high-lift device optimization. Comput Fluids2016; 124: 105–116.
3.
MoensF.DandoisJ. Optimization of passive flow control devices of a slatless high-lift configuration. J Aircraft2016; 53(1): 189–201.
4.
WangXWangSTaoJ, et al.A PCA-ANN-based inverse design model of stall lift robustness for high-lift device. Aerospace Sci Tech2018; 81: 272–283.
5.
TangZPériauxJBugedaG, et al.Lift maximization with uncertainties for the optimization of high-lift devices. Int J Numer Meth Fluids2010; 64(2): 119–135.
6.
TaoJSunG. An artificial neural network approach for aerodynamic performance retention in airframe noise reduction design of a 3D swept wing model. Chin J Aeronautics2016; 29(5): 1213–1225.
7.
HerkertJBorensteinJMillerK. The Boeing 737 MAX: Lessons for engineering ethics. Sci Eng Ethics2020; 26(6): 2957–2974.
8.
NelsonRC. Flight Stability and Automatic control. NY, USA: WCB/McGraw Hill, 1998.
9.
EstebanS. Static and dynamic analysis of an unconventional plane-flying wing. In: AIAA atmospheric flight mechanics conference and exhibit - guidance, navigation, and control and co-located conferences, Montreal, Canada, 06 August 2001.
10.
NgTTHLengGSB. Application of genetic algorithms to conceptual design of a micro-air vehicle. Eng Appl Artif Intelligence2002; 15(5): 439–445.
11.
ReistTAZinggDWRakowitzM, et al.Multifidelity optimization of hybrid wing-body aircraft with stability and control requirements. J Aircraft2019; 56(2): 442–456.
12.
BakCFuglsangPJohansenJ, et al.Wind tunnel tests of the NACA 63-415 and a modified NACA 63-415 airfoil, 2000.
13.
GopinathAJamesonA. Time spectral method for periodic unsteady computations over two-and three-dimensional bodies[C]. In: 43rd AIAA aerospace sciences meeting and exhibit, Reno, Nevada, 10 January 2005.
14.
BhagwandinVASahuJ. Numerical prediction of pitch damping stability derivatives for finned projectiles. J Spacecraft Rockets2014; 51(5): 1603–1618.
15.
EsfahaniVNManiM. Numerical study of reduced frequency effect on longitudinal stability derivatives of airfoil under pitching and plunging oscillations. J.Aerosp Technol Manag2016; 8(3): 272–280.
16.
HanZ-HGörtzS. Hierarchical kriging model for variable-fidelity surrogate modeling. AIAA J2012; 50(9): 1885–1896.
17.
HanZ-HZimmermannRGörtzS. Alternative cokriging method for variable-fidelity surrogate modeling. AIAA J2012; 50(5): 1205–1210.
18.
HanZZhangK. Surrogate-based optimization. Real-world applications of genetic algorithms. IntechOpen, 2012, 343, pp. 343–345.
19.
SunG.WangS. A review of the artificial neural network surrogate modeling in aerodynamic design. Proc Inst Mech Eng G: J Aerospace Eng2019; 233(16): 5863–5872.
20.
LecunYBengioYHintonG. Deep learning. Nature2015; 521(7553): 436–444.
21.
SekarVZhangMShuC, et al.Inverse design of airfoil using a deep convolutional neural network. Aiaa J2019; 57(3): 993–1003.
22.
WangSSunGChenW, et al.Database self-expansion based on artificial neural network: An approach in aircraft design. Aerospace Sci Tech2018; 72: 77–83.
23.
ChangCFangJDinhNT. Reynolds-averaged turbulence modeling using deep learning with local flow features: An empirical approach. Nucl Sci Eng2020; 194(8–9): 650–664.
24.
LingJKurzawskiATempletonJ. Reynolds averaged turbulence modelling using deep neural networks with embedded invariance. J Fluid Mech2016; 807: 155–166.
TatarMMasdariM. Investigation of pitch damping derivatives for the standard dynamic model at high angles of attack using neural network. Aerospace Sci Tech2019; 92: 685–695.
27.
WinterMBreitsamterC. Nonlinear identification via connected neural networks for unsteady aerodynamic analysis. Aerospace Sci Tech2018; 77: 802–818.
28.
GersFASchmidhuberJCumminsF. Learning to forget: Continual prediction with LSTM. In: 1999 ninth international conference on artificial neural networks ICANN 99, Edinburgh, UK, 7–10 September 1999. IET.
29.
HochreiterSSchmidhuberJ. Long short-term memory. Neural Comput1997; 9(8): 1735–1780.
30.
MohanATGaitondeDV. A deep learning based approach to reduced order modeling for turbulent flow control using LSTM neural networks, 2018.arXiv preprint arXiv:1804.09269
31.
LiKKouJZhangW. Deep neural network for unsteady aerodynamic and aeroelastic modeling across multiple Mach numbers. Nonlinear Dyn2019; 96(3): 2157–2177.
32.
KouJZhangW. An approach to enhance the generalization capability of nonlinear aerodynamic reduced-order models. Aerospace Sci Technol2016; 49: 197–208.
33.
PengLPengMLiaoB, et al.The advances and challenges of deep learning application in biological big data processing. Cbio2018; 13(4): 352–359.
34.
LeeTGerontakosP. Investigation of flow over an oscillating airfoil. J Fluid Mech2004; 512: 313.
35.
GharaliKJohnsonDA. Dynamic stall simulation of a pitching airfoil under unsteady freestream velocity. J Fluids Structures2013; 42: 228–244.
36.
WangSInghamDBMaL, et al.Turbulence modeling of deep dynamic stall at relatively low Reynolds number. J Fluids Structures2012; 33: 191–209.
37.
MukherjeeSRoyS. Enhancement of lift and drag characteristics of an oscillating airfoil in deep dynamic stall using plasma actuation. In: ,50th AIAA aerospace sciences meeting including the new horizons forum and aerospace expositionNashville, Tennessee, 09 January 2012.
38.
CampbellJ PPaulsonJ W. The Effects of Static Margin and Rotational Damping in Pitch on the Longitudinal Stability Characteristics of an Airplane as Determined by Tests of a Model in the NACA. National Aeronautics And Space Admin Langley Research Center Hampton Va, 1944.
39.
D’SouzaRNHuangPYehF. Structural analysis and optimization of convolutional neural networks with a small sample size. Scientific Rep2020; 10(1): 1–13.
40.
BoutemedjetASamardžićMRebhiL, et al.UAV aerodynamic design involving genetic algorithm and artificial neural network for wing preliminary computation. Aerospace Sci Tech2019; 84: 464–483.
41.
HaryantoIUtomoT SSinagaN, et al.Optimization of maximum lift to drag ratio on airfoil design based on artificial neural network utilizing genetic algorithm. Appli Mech Material2014; 493: 123–128.
EberhartRKennedyJ. A new optimizer using particle swarm theory. In: ,MHS’95. Proceedings of the sixth international symposium on micro machine and human scienceNagoya, Japan, 4–6 October 1995. IEEE.
