Sparse Pressure-Based Machine Learning Approach for Aerodynamic Loads Estimation During Gust Encounters
Abstract
Estimation of aerodynamic loads is a significant challenge in complex gusty environments due to the associated complexities of flow separation and strong nonlinearities. In this study, we explore the practical feasibility of multilayer perceptron (MLP) for estimating aerodynamic loads in gusts, when confounded by noisy and spatially distributed sparse surface pressure measurements. As a demonstration, a nonslender delta wing experiencing various gusts with different initial and final conditions is considered. Time-resolved lift and drag, and spatially distributed sparse surface pressure measurements are collected in a towing-tank facility. The nonlinear MLP model is used to estimate gust scenarios that are unseen in training progress. A filtering process allows us to examine the fluctuation of the dynamic response from the pressure measurements on the MLP. Estimation results show that the MLP model is able to capture the relationship between surface pressure and aerodynamic loads with a minimum quantity of learning samples, delivering accurate estimations, despite the slightly large errors for the cases at the boundary of the datasets. The results also indicate that the dynamic response of the pressure measurements has an influence on the learning of MLP. We further utilize gradient maps to perform a sensitivity analysis, so as to evaluate the contribution of the pressure data to the estimation of gust loads. This study reveals the significant contribution of the sensors located near the leading edge and at the nose of the delta wing. Our findings suggest the potential for an efficient sensor deployment strategy in data-driven aerodynamic load estimation.
References
[1] , “Recent Studies of Subsonic Vortex Lift Including Parameters Affecting Stable Leading-Edge Vortex Flow,” Journal of Aircraft, Vol. 14, No. 12, 1977, pp. 1205–1211. https://doi.org/10.2514/3.58916
[2] , “Review of the Physics of Enhancing Vortex Lift by Unsteady Excitation,” Progress in Aerospace Sciences, Vol. 28, No. 2, 1991, pp. 73–131. https://doi.org/10.1016/0376-0421(91)90001-K
[3] , “Low Speed Aerodynamic Performance Analysis of Vortex Lift Waveriders with a Wide-Speed Range,” Acta Astronautica, Vol. 161, Aug. 2019, pp. 209–221. https://doi.org/10.1016/j.actaastro.2019.05.029
[4] , “A Numerical Method for the Calculation of Nonlinear, Unsteady Lifting Potential Flow Problems,” AIAA Journal, Vol. 7, No. 10, 1969, pp. 2001–2009. https://doi.org/10.2514/3.5494
[5] , “State-Space Modeling for Aeroelastic Panels with Linearized Potential Flow Aerodynamic Loading,” Journal of Aircraft, Vol. 33, No. 4, 1996, pp. 816–822. https://doi.org/10.2514/3.47019
[6] , “Comparing Different Potential Flow Methods for Unsteady Aerodynamic Modelling of a Flutter Demonstrator Aircraft,” AIAA SciTech 2023 Forum, AIAA Paper 2023-0177, 2023. https://doi.org/10.2514/6.2023-0177
[7] , “A Concept of the Vortex Lift of Sharp-Edge Delta Wings Based on a Leading-Edge-Suction Analogy,” NASA TN d-3767, 1966.
[8] , “Parameter-Varying Aerodynamics Models for Aggressive Pitching-Response Prediction,” AIAA Journal, Vol. 55, No. 3, 2017, pp. 693–701. https://doi.org/10.2514/1.J055193
[9] , “Hydrodynamic Stimuli and the Fish Lateral Line,” Nature, Vol. 408, No. 6808, 2000, pp. 51–52. https://doi.org/10.1038/35040706
[10] , “Sensory Ecology of the Fish Lateral-Line System: Morphological and Physiological Adaptations for the Perception of Hydrodynamic Stimuli,” Journal of Fish Biology, Vol. 95, No. 1, 2019, pp. 53–72. https://doi.org/10.1111/jfb.13966
[11] , “Bird Wings Act as a Suspension System that Rejects Gusts,” Proceedings of the Royal Society B, Vol. 287, No. 1937, 2020, Paper 2020-1748. https://doi.org/10.1098/rspb.2020.1748
[12] , “Theoretical Morphology and Development of Flight Feather Vane Asymmetry with Experimental Tests in Parrots,” Journal of Experimental Zoology Part B: Molecular and Developmental Evolution, Vol. 322, No. 4, 2014, pp. 240–255. https://doi.org/10.1002/jez.b.22573
[13] , Machine Learning Control-Taming Nonlinear Dynamics and Turbulence, Vol. 116, Springer, Berlin, 2017, pp. 1–10.
[14] , Data-Driven Science and Engineering: Machine Learning, Dynamical Systems, and Control, Cambridge Univ. Press, Cambridge, MA, 2019, pp. 264–274.
[15] , “Assessment of Supervised Machine Learning Methods for Fluid Flows,” Theoretical and Computational Fluid Dynamics, Vol. 34, No. 4, 2020, pp. 497–519. https://doi.org/10.1007/s00162-020-00518-y
[16] , “Machine Learning for Fluid Mechanics,” Annual Review of Fluid Mechanics, Vol. 52, No. 1, 2020, pp. 477–508. https://doi.org/10.1146/annurev-fluid-010719-060214
[17] , “Machine-Learning-Based Spatio-Temporal Super Resolution Reconstruction of Turbulent Flows,” Journal of Fluid Mechanics, Vol. 909, Feb. 2021, p. A9. https://doi.org/10.1017/jfm.2020.948
[18] , “Unsupervised Deep Learning for Super-Resolution Reconstruction of Turbulence,” Journal of Fluid Mechanics, Vol. 910, March 2021, p. A29. https://doi.org/10.1017/jfm.2020.1028
[19] , “Special Issue on Machine Learning and Data-Driven Methods in Fluid Dynamics,” Theoretical and Computational Fluid Dynamics, Vol. 34, No. 4, 2020, pp. 333–337. https://doi.org/10.1007/s00162-020-00542-y
[20] , “Perspective on Machine Learning for Advancing Fluid Mechanics,” Physical Review Fluids, Vol. 4, No. 10, 2019, Paper 100501. https://doi.org/10.1103/PhysRevFluids.4.100501
[21] , “Flowfield Reconstruction Method Using Artificial Neural Network,” AIAA Journal, Vol. 57, No. 2, 2019, pp. 482–498. https://doi.org/10.2514/1.J057108
[22] , “Shallow Neural Networks for Fluid Flow Reconstruction with Limited Sensors,” Proceedings of the Royal Society A, Vol. 476, No. 2238, 2020, Paper 2020-0097. https://doi.org/10.1098/rspa.2020.0097
[23] , “Using Machine Learning to Detect the Turbulent Region in Flow Past a Circular Cylinder,” Journal of Fluid Mechanics, Vol. 905, Dec. 2020, p. A10. https://doi.org/10.1017/jfm.2020.725
[24] , “Machine Learning to Classify Vortex Wakes of Energy Harvesting Oscillating Foils,” AIAA Journal, Vol. 61, No. 3, 2023, pp. 1281–1291. https://doi.org/10.2514/1.J062091
[25] , “Leading Edge Vortex Formation and Wake Trajectory: Synthesizing Measurements, Analysis, and Machine Learning,” Physical Review Fluids, Vol. 7, No. 7, 2022, Paper 074704. https://doi.org/10.1103/PhysRevFluids.7.074704
[26] , “Learning Representations by Back-Propagating Errors,” Nature, Vol. 323, No. 6088, 1986, pp. 533–536. https://doi.org/10.1038/323533a0
[27] , “Identification of Aerodynamic Coefficients Using Computational Neural Networks,” Journal of Guidance, Control, and Dynamics, Vol. 16, No. 6, 1993, pp. 1018–1025. https://doi.org/10.2514/3.21122
[28] , “Neural Network Prediction of Three-Dimensional Unsteady Separated Flowfields,” Journal of Aircraft, Vol. 32, No. 1, 1995, pp. 178–185. https://doi.org/10.2514/3.46698
[29] , “Neural Networks: Applications and Opportunities in Aeronautics,” Progress in Aerospace Sciences, Vol. 32, No. 5, 1996, pp. 433–456. https://doi.org/10.1016/0376-0421(95)00011-9
[30] , “Reynolds Averaged Turbulence Modelling Using Deep Neural Networks with Embedded Invariance,” Journal of Fluid Mechanics, Vol. 807, Nov. 2016, pp. 155–166. https://doi.org/10.1017/jfm.2016.615
[31] , “Construction of Reduced-Order Models for Fluid Flows Using Deep Feedforward Neural Networks,” Journal of Fluid Mechanics, Vol. 872, Aug. 2019, pp. 963–994. https://doi.org/10.1017/jfm.2019.358
[32] , “Exploring the Signature of Distributed Pressure Measurements on Non-Slender Delta Wings During Axial and Vertical Gusts,” Physics of Fluids, Vol. 32, No. 11, 2020, Paper 115110. https://doi.org/10.1063/5.0025860
[33] , “Load Estimation in Unsteady Flows from Sparse Pressure Measurements: Application of Transition Networks to Experimental Data,” Physics of Fluids, Vol. 34, No. 2, 2022, Paper 025105. https://doi.org/10.1063/5.0076731
[34] , “Pressure Feedback Control of Aerodynamic Loads on a Delta Wing in Transverse Gusts,” AIAA Journal, Vol. 61, No. 4, 2023, pp. 1659–1674. https://doi.org/10.2514/1.J062442
[35] , “Random Forests,” Machine Learning, Vol. 45, No. 1, 2001, pp. 5–32. https://doi.org/10.1023/A:1010933404324
[36] , “A Tutorial on Support Vector Regression,” Statistics and Computing, Vol. 14, No. 3, 2004, pp. 199–222. https://doi.org/10.1023/B:STCO.0000035301.49549.88
[37] , “A Learning Algorithm for Continually Running Fully Recurrent Neural Networks,” Neural Computation, Vol. 1, No. 2, 1989, pp. 270–280. https://doi.org/10.1162/neco.1989.1.2.270
[38] , “Gradient-Based Learning Applied to Document Recognition,” Proceedings of the IEEE, Vol. 86, No. 11, 1998, pp. 2278–2324. https://doi.org/10.1109/5.726791
[39] , “Random Forest Regression-Based Machine Learning Model for Accurate Estimation of Fluid Flow in Curved Pipes,” Processes, Vol. 9, No. 11, 2021, p. 2095. https://doi.org/10.3390/pr9112095
[40] , “Unsteady Aerodynamic Modeling at High Angles of Attack Using Support Vector Machines,” Chinese Journal of Aeronautics, Vol. 28, No. 3, 2015, pp. 659–668. https://doi.org/10.1016/j.cja.2015.03.010
[41] , “Time-Resolved Turbulent Velocity Field Reconstruction Using a Long Short-Term Memory (LSTM)-Based Artificial Intelligence Framework,” Physics of Fluids, Vol. 31, No. 7, 2019, Paper 075108. https://doi.org/10.1063/1.5111558
[42] , “Unsteady Aerodynamic Modeling Based on Recurrent Neural Network,” Aerodynamic Research & Experiment, Vol. 32, No. 1, 2020, p. 101. https://doi.org/10.12050/are20200107
[43] , “Physics and Modeling of Large Flow Disturbances: Discrete Gust Encounters for Modern Air Vehicles,” Annual Review of Fluid Mechanics, Vol. 54, No. 1, 2022, pp. 469–493. https://doi.org/10.1146/annurev-fluid-031621-085520
[44] , “Phenomena-Based Real-Time Aerodynamic Measurement System (PRAMS),” 2003 IEEE Aerospace Conference Proceedings (Cat. No. 03TH8652), Vol. 7, IEEE, New York, 2003, pp. 3347–3356. https://doi.org/10.1109/AERO.2003.1234177
[45] , “Unsteady Aerodynamic Observables for Gust Load Alleviation,” 49th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference, 16th AIAA/ASME/AHS Adaptive Structures Conference, 10th AIAA Non-Deterministic Approaches Conference, 9th AIAA Gossamer Spacecraft Forum, 4th AIAA Multidisciplinary Design Optimization Specialists Conference, AIAA Paper 2008-1725, 2008. https://doi.org/10.2514/6.2008-1725
[46] , Gust Loads on Aircraft: Concepts and Applications,
AIAA Education Series , AIAA, Reston, VA, 1988, pp. 7–20.[47] , “Aerodynamic Parameters from Distributed Heterogeneous CNT Hair Sensors with a Feedforward Neural Network,” Bioinspiration & Biomimetics, Vol. 11, No. 6, 2016, Paper 066006. https://doi.org/10.1088/1748-3190/11/6/066006
[48] , “Neural-Network-Based Flush Air Data Sensing System Demonstrated on a Mini Air Vehicle,” Journal of Aircraft, Vol. 47, No. 1, 2010, pp. 18–31. https://doi.org/10.2514/1.44157
[49] , “Pitch Control of a Micro Air Vehicle with Micropressure Sensors,” Journal of Aircraft, Vol. 50, No. 1, 2013, pp. 239–248. https://doi.org/10.2514/1.C031894
[50] , “Aerodynamic Moment Model Calibration from Distributed Pressure Arrays,” Journal of Aircraft, Vol. 54, No. 2, 2017, pp. 716–723. https://doi.org/10.2514/1.C033898
[51] , “Machine-Learning-Based Detection of Aerodynamic Disturbances Using Surface Pressure Measurements,” AIAA Journal, Vol. 57, No. 12, 2019, pp. 5079–5093. https://doi.org/10.2514/1.J058486
[52] , “Distributed Pressure Sensing–Based Flight Control for Small Fixed-Wing Unmanned Aerial Systems,” Journal of Aircraft, Vol. 56, No. 5, 2019, pp. 1951–1960. https://doi.org/10.2514/1.C035416
[53] , “Sparse Sensor Reconstruction of Vortex-Impinged Airfoil Wake with Machine Learning,” Theoretical and Computational Fluid Dynamics, Vol. 37, April 2023, pp. 269–287. https://doi.org/10.1007/s00162-023-00657-y
[54] , “Characterization of and Correction for Pressure-Measurement Installation,” Trita-Mek Technical Report, Centres, Competence Center for Gas Exchange (CCGEx), KTH Mechanics, Royal Institute of Technology, Stockholm, Sweden, 2017, p. 11.
[55] , “Dynamics of Pneumatic Transmission Lines,” Control Engineering, Vol. 10, No. 9, 1963, pp. 114–117.
[56] , Theoretical and Experimental Results for the Dynamic Response of Pressure Measuring Systems, Rept. NLR-TR F.238, National Lucht- en Ruimtevaartlaboratorium, Amsterdam, Netherlands, 1965.
[57] , “Correction of Distortion Effects Caused by Tubing Systems in Measurements of Fluctuating Pressures,” Journal of Wind Engineering and Industrial Aerodynamics, Vol. 5, Nos. 1–2, 1979, pp. 93–107. https://doi.org/10.1016/0167-6105(79)90026-6
[58] , “Experimental and Theoretical Frequency Response of Pressure Transducers for High Speed Turbomachinery,” International Journal of Turbo and Jet Engines, Vol. 17, No. 2, 2000, pp. 153–160. https://doi.org/10.1515/TJJ.2000.17.2.153
[59] , “Wiener Deconvolution for Reconstruction of Pneumatically Attenuated Pressure Signals,” AIAA Journal, Vol. 49, No. 5, 2011, pp. 890–897. https://doi.org/10.2514/1.J050102
[60] , “Pressure Correction Schemes and the Use of the Wiener Deconvolution Method in Pneumatic Systems with Short Tubes,” Experiments in Fluids, Vol. 53, June 2012, pp. 829–837. https://doi.org/10.1007/s00348-012-1332-2
[61] , “Characterization and Control of Unsteady Aerodynamics on Wind Turbine Aerofoils,” Journal of Physics: Conference Series, Vol. 524, No. 1, June 2014, Paper 012025. https://doi.org/10.1088/1742-6596/524/1/012025
[62] , “Quantification of Uncertainty in the Correction of Remotely Measured Unsteady Pressure Signals on Pitching Airfoils,” 33rd AIAA Aerodynamic Measurement Technology and Ground Testing Conference, AIAA Paper 2017-3733, 2017. https://doi.org/10.2514/6.2017-3733
[63] , “Analysis of the Unsteady Surface Pressure Distribution of a Pitching Airfoil Using Modal Decomposition,” Experiments in Fluids, Vol. 64, No. 5, 2023, pp. 1–19. https://doi.org/10.1007/s00348-023-03644-5
[64] , “Measurement of the Volume Displacement of Blood-Pressure Transducers,” Medical and Biological Engineering and Computing, Vol. 22, No. 6, 1984, pp. 613–614. https://doi.org/10.1007/BF02443881
[65] , “An Attempt to Measure Fluctuating Local Pressure in Free Turbulent Flow in Water,” Journal of Fluid Science and Technology, Vol. 9, No. 2, 2014, Paper JFST0014.https://doi.org/10.1299/jfst.2014jfst0014
[66] , “Grad-CAM: Why Did you Say That?” arXiv preprint, arXiv:1611.07450, 2016. https://doi.org/10.48550/arXiv.1611.07450
[67] , “Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization,” Proceedings of the IEEE International Conference on Computer Vision, 2017, pp. 618–626.
[68] , “Inverse Identification of Dynamically Important Regions in Turbulent Flows Using Three-Dimensional Convolutional Neural Networks,” Physical Review Fluids, Vol. 8, No. 9, Sept. 2023, Paper 094605. https://doi.org/10.1103/PhysRevFluids.8.094605
[69] , “Interpretable Deep Learning for Prediction of Prandtl Number Effect in Turbulent Heat Transfer,” Journal of Fluid Mechanics, Vol. 955, Jan. 2023, p. A14. https://doi.org/10.1017/jfm.2022.1069
[70] , “Generalization Techniques of Neural Networks for Fluid Flow Estimation,” Neural Computing and Applications, Vol. 34, No. 5, 2022, pp. 3647–3669. https://doi.org/10.1007/s00521-021-06633-z
[71] , “Data-Driven Sparse Sensor Placement for Reconstruction: Demonstrating the Benefits of Exploiting Known Patterns,” IEEE Control Systems Magazine, Vol. 38, No. 3, 2018, pp. 63–86. https://doi.org/10.1109/MCS.2018.2810460
[72] , “Greedy Sensor Placement with Cost Constraints,” IEEE Sensors Journal, Vol. 19, No. 7, 2018, pp. 2642–2656. https://doi.org/10.1109/JSEN.2018.2887044
[73] , “Determinant-Based Fast Greedy Sensor Selection Algorithm,” IEEE Access, Vol. 9, April 2021, pp. 68,535–68,551. https://doi.org/10.1109/ACCESS.2021.3076186
[74] , “Fast Greedy Optimization of Sensor Selection in Measurement with Correlated Noise,” Mechanical Systems and Signal Processing, Vol. 158, Sept. 2021, Paper 107619. https://doi.org/10.1016/j.ymssp.2021.107619
[75] , “Separation Mechanics of Non-Slender Delta Wings During Streamwise Gusts,” Journal of Fluids and Structures, Vol. 90, Oct. 2019, pp. 286–296. https://doi.org/10.1016/j.jfluidstructs.2019.07.001
[76] , “Delving Deep into Rectifiers: Surpassing Human-Level Performance on Imagenet Classification,” Proceedings of the IEEE International Conference on Computer Vision, 2015, pp. 1026–1034.
[77] , “Adam: A Method for Stochastic Optimization,” arXiv preprint, arXiv:1412.6980, 2014. https://doi.org/10.48550/arXiv.1412.6980
[78] , “Tensorflow: Large-Scale Machine Learning on Heterogeneous Distributed Systems,” arXiv preprint, arXiv:1603.04467, 2016. https://doi.org/10.48550/arXiv.1603.04467
[79] , “Keras: The Python Deep Learning Library,” Astrophysics Source Code Library, 2018, pp. ascl.1806.
[80] “Talos [Computer Software],” Autonomio, 2020, http://github.com/autonomio/talos.
[81] , “Random Search for Hyper-Parameter Optimization,” Journal of Machine Learning Research, Vol. 13, No. 2, 2012, pp. 281–305.
[82] , “Practical Bayesian Optimization of Machine Learning Algorithms,” 26th Annual Conference on Neural Information Processing Systems, Vol. 4, NIPS, Lake Tahoe, NV, 2012, pp. 2951–2959.