APTNet: A Condition-Sensitive Modulation Framework for Surface Pressure Prediction on Supersonic Aircraft

Hongbin Xu , Yin Long , Junlin Wu

CAAI Transactions on Intelligence Technology ›› 2026, Vol. 11 ›› Issue (4) : 1302 -1319.

PDF (4165KB)
CAAI Transactions on Intelligence Technology ›› 2026, Vol. 11 ›› Issue (4) :1302 -1319. DOI: 10.1049/cit2.70149
ORIGINAL RESEARCH
research-article
APTNet: A Condition-Sensitive Modulation Framework for Surface Pressure Prediction on Supersonic Aircraft
Author information +
History +
PDF (4165KB)

Abstract

Accurate surface-pressure prediction over broad operating envelopes is critical for supersonic aerodynamic analysis and design. To overcome the bottlenecks of traditional computational fluid dynamics (CFD) in real-time performance and computational efficiency, data-driven deep learning methods have emerged. However, existing data-driven models face two fundamental physical challenges: first, the pressure response to Mach number, angle of attack and altitude is strongly spatially heterogeneous, which cannot be adequately captured by traditional global condition fusion methods; second, the inherent strong anisotropy of supersonic flows renders standard isotropic graph construction based on Euclidean distance prone to spurious cross-shock connections, resulting in nonphysical smoothing of predictions. To address these issues, we propose aircraft pressure transformer network (APTNet), a physics-informed deep learning framework that explicitly encodes flow–geometry interactions and streamwise physical priors. Specifically, we first propose a condition-sensitive local modulation (CSLM) module, which utilises a surface-normal–freestream alignment term to distinguish windward and leeward regions, and learns per-point sensitivity fields to implement physically interpretable local pressure modulation via a feature-wise linear modulation (FiLM) mechanism. Furthermore, we introduce a flow-aligned anisotropic graph construction strategy, which reshapes the local receptive fields of graph nodes into ellipsoids elongated along streamlines, enforcing information to propagate preferentially along streamlines and suppressing cross-shock feature mixing. Experiments on diverse complex aircraft configurations demonstrate that APTNet consistently outperforms existing baseline models in terms of both accuracy and efficiency. Beyond quantitative improvements, the model provides interpretable intermediate signals, such as attention maps and feature responses, which can be directly integrated into the aerodynamic design workflow.

Keywords

aerodynamic prediction / condition-sensitive local modulation / flow-aligned graph attention / physics-informed deep learning / surface pressure prediction

Cite this article

Download citation ▾
Hongbin Xu, Yin Long, Junlin Wu. APTNet: A Condition-Sensitive Modulation Framework for Surface Pressure Prediction on Supersonic Aircraft. CAAI Transactions on Intelligence Technology, 2026, 11 (4) : 1302-1319 DOI:10.1049/cit2.70149

登录浏览全文

4963

注册一个新账户 忘记密码

References

[1]

N. Thuerey, K. Weißenow, L. Prantl, and X. Hu, “ Deep Learning Methods for Reynolds-Averaged Navier-Stokes Simulations of Airfoil Flows,” AIAA Journal 58, no. 1 (2020): 25-36, https://doi.org/10.2514/1.j058291.

[2]

R. Vinuesa and S. L. Brunton, “ Enhancing Computational Fluid Dynamics With Machine Learning,” Nature Computational Science 2, no. 6 (2022): 358-366, https://doi.org/10.1038/s43588-022-00264-7.

[3]

Jh Lin, Ss Chen, Jp Li, et al., “ A Velocity-Prediction Lightweight Diffusion Model for Three-Dimensional Aircraft Aerodynamic Inverse Design,” Engineering Applications of Artificial Intelligence 163 (2026): 113125, https://doi.org/10.1016/j.engappai.2025.113125.

[4]

Li Yg, Ss Chen, Sy Jin, Jh Lin, Ml Jia, and Qf Jiang, “ Mixed Conditional Generative Adversarial Networks for High-Dimensional Aircraft Aerodynamic Stealth Shape Optimization,” Engineering Applications of Artificial Intelligence 173 (2026): 114483, https://doi.org/10.1016/j.engappai.2026.114483.

[5]

Jh Lin, Ss Chen, Sy Jin, Qf Jiang, Ml Jia, and D. Li, “ Multi-Point Aerodynamic Inverse Design of Flying Wing Using the Conditional Diffusion Model,” Physics of Fluids 37, no. 7 (2025): 077116, https://doi.org/10.1063/5.0271723.

[6]

T. Pfaff, M. Fortunato, A. Sanchez-Gonzalez, and P. Battaglia, Learning Mesh-Based Simulation With Graph Networks (OpenReview.net, 2020).

[7]

D. Roznowicz, G. Stabile, N. Demo, D. Fransos, and G. Rozza, “ Large-Scale Graph-Machine-Learning Surrogate Models for 3D-Flowfield Prediction in External Aerodynamics,” Advanced Modeling and Simulation in Engineering Sciences 11, no. 1 (2024): 6, https://doi.org/10.1186/s40323-024-00259-1.

[8]

R. F. Miotto and W. R. Wolf, “ Flow Imaging as an Alternative to Non-Intrusive Measurements and Surrogate Models Through Vision Transformers and Convolutional Neural Networks,” Physics of Fluids 35, no. 4 (2023): 045143, https://doi.org/10.1063/5.0144700.

[9]

X. Wang, S. Zhu, Y. Guo, et al., “ TransFlowNet: A Physics-Constrained Transformer Framework For Spatio-Temporal Super-Resolution of Flow Simulations,” Journal of Computational Science 65 (2022): 101906, https://doi.org/10.1016/j.jocs.2022.101906.

[10]

X. Guo, W. Li, and F. Iorio, “ Convolutional Neural Networks for Steady Flow Approximation,” in Proceedings of the 22nd ACM SIGKDD International Conference, (2016).

[11]

V. Sekar, Q. Jiang, C. Shu, and B. C. Khoo, “ Fast Flow Field Prediction Over Airfoils Using Deep Learning Approach,” Physics of Fluids 31, no. 5 (2019): 057103, https://doi.org/10.1063/1.5094943.

[12]

X. Hui, J. Bai, H. Wang, and Y. Zhang, “ Fast Pressure Distribution Prediction of Airfoils Using Deep Learning,” Aerospace Science and Technology 105 (2020): 105949, https://doi.org/10.1016/j.ast.2020.105949.

[13]

S. L. Brunton, B. R. Noack, and P. Koumoutsakos, “ Machine Learning for Fluid Mechanics,” Annual Review of Fluid Mechanics 52, no. 1 (2020): 477-508, https://doi.org/10.1146/annurev-fluid-010719-060214.

[14]

K. Duraisamy, G. Iaccarino, and H. Xiao, “ Turbulence Modeling in the Age of Data,” Annual Review of Fluid Mechanics 51, no. 1 (2019): 357-377, https://doi.org/10.1146/annurev-fluid-010518-040547.

[15]

Y. Shen, W. Huang, Zg Wang, Df Xu, and C. Y. Liu, “ A Deep Learning Framework for Aerodynamic Pressure Prediction on General Three-Dimensional Configurations,” Physics of Fluids 35, no. 10 (2023): 107111, https://doi.org/10.1063/5.0172437.

[16]

F. Xiong, L. Zhang, X. Hu, and C. Ren, “ A Point Cloud Deep Neural Network Metamodel Method for Aerodynamic Prediction,” Chinese Journal of Aeronautics 36, no. 4 (2023): 92-103, https://doi.org/10.1016/j.cja.2022.11.025.

[17]

S. Wang, X. Zhang, Y. Tie, W. Xia, and Y. Jiang, “ Aerodynamic Performance Prediction of 3D Aircraft Based on Point Cloud Deep Learning,” Journal of Computational Design and Engineering 12, no. 5 (2025): 175-190, https://doi.org/10.1093/jcde/qwaf051.

[18]

M. Elrefaie, A. Dai, and F. Ahmed, “ DrivAerNet: A Parametric Car Dataset for Data-Driven Aerodynamic Design and Prediction,” Journal of Mechanical Design 147, no. 4 (2025): 041712, https://doi.org/10.1115/1.4068104.

[19]

H. Zhang, Y. Shen, W. Huang, Z. Xie, and Yb Niu, “ Deep Transfer Learning For Three-Dimensional Aerodynamic Pressure Prediction Under Data Scarcity,” Theoretical and Applied Mechanics Letters 15, no. 2 (2025): 100571, https://doi.org/10.1016/j.taml.2025.100571.

[20]

Z. Zhu, G. Zhao, and Q. Zhao, “ Fast and High-Precision Compressible Flowfield Inference Method of Transonic Airfoils Based on Attention Unet,” Physics of Fluids 36, no. 3 (2024): 036111, https://doi.org/10.1063/5.0188550.

[21]

K. Zuo, Z. Ye, W. Zhang, X. Yuan, and L. Zhu, “ Fast Aerodynamics Prediction of Laminar Airfoils Based on Deep Attention Network,” Physics of Fluids 35, no. 3 (2023): 037127, https://doi.org/10.1063/5.0140545.

[22]

M. M. Bronstein, J. Bruna, T. Cohen, and P. Veličković, “ Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges,” arXiv preprint arXiv:2104.13478 (2021).

[23]

P. W. Battaglia, J. B. Hamrick, V. Bapst, et al., “ Relational Inductive Biases, Deep Learning, and Graph Networks,” arXiv preprint arXiv:1806.01261 (2018).

[24]

A. Sanchez-Gonzalez, J. Godwin, T. Pfaff, R. Ying, J. Leskovec, and P. Battaglia, Learning to Simulate Complex Physics With Graph Networks (PMLR, 2020), 8459-8468.

[25]

F. D. A. Belbute-Peres, T. Economon, and Z. Kolter, Combining Differentiable PDE Solvers and Graph Neural Networks for Fluid Flow Prediction (PMLR, 2020), 2402-2411.

[26]

D. Hines and P. Bekemeyer, “ Graph Neural Networks for the Prediction of Aircraft Surface Pressure Distributions,” Aerospace Science and Technology 137 (2023): 108268, https://doi.org/10.1016/j.ast.2023.108268.

[27]

C. Ouyang, H. Liu, T. Ma, S. Xu, and X. Xi, “ A Physics-Integrated Attention Model With Dynamic Weighting for Rapid and Precise Turbulent Flow Prediction in Variable Geometries,” Physics of Fluids 37, no. 8 (2025): 085103, https://doi.org/10.1063/5.0278647.

[28]

A. Hemmasian and F. A. Barati, “ Reduced-Order Modeling of Fluid Flows With Transformers,” Physics of Fluids 35, no. 5 (2023), https://doi.org/10.1063/5.0151515.

[29]

Q. Cui, M. Zhang, M. Xiao, and G. Ni, “ Transformer Based Deep Learning Accelerated Numerical Simulation for Incompressible Flow,” Physics of Fluids 36, no. 12 (2024): 127152, https://doi.org/10.1063/5.0243174.

[30]

X. Hu, J. Zhang, K. Yan, T. Wan, and X. Zheng, Physics-Informed Transformer for Efficient Fluid Dynamics Predictions (Springer, 2025), 356-368.

[31]

Y. Fan, C. Wei, J. C. Wong, C. C. Ooi, H. Wang, and P. H. Chiu, “ Cascaded PhyFormer-SRNet: A Physics-Informed Transformer-Based Framework With Cascaded Refinement for Spatiotemporal Super-Resolution of Turbulent Flows,” Physics of Fluids 37, no. 7 (2025): 075207, https://doi.org/10.1063/5.0274979.

[32]

E. Perez, F. Strub, H. De Vries, V. Dumoulin, and A. Courville, “ FiLM: Visual Reasoning With a General Conditioning Layer,” in AAAI’18/IAAI’18/EAAI’18 (AAAI Press, 2018).

[33]

H. De Vries, F. Strub, J. Mary, H. Larochelle, O. Pietquin, and A. C. Courville, “ Modulating Early Visual Processing by Language,” in Advances in Neural Information Processing Systems, vol. 30 (NeurIPS, 2017).

[34]

T. Li, L. Guo, Z. Yang, et al., “ An Automatic Shape-Aware Method for Predicting Heat Flux of Supersonic Aircraft Based on a Deep Learning Approach,” Physics of Fluids 34, no. 7 (2022): 077103, https://doi.org/10.1063/5.0098341.

PDF (4165KB)

0

Accesses

0

Citation

Detail

Sections
Recommended

/

〈 〉