Kolmogorov–Arnold Networks for Reduced-Order Modeling in Unsteady Aerodynamics and Aeroelasticity
Source abstract
Abstract
Kolmogorov–Arnold Networks (KANs) are a recent development in machine learning, offering strong functional representation capabilities, enhanced interpretability, and reduced parameter complexity. Leveraging these advantages, this paper proposes a KAN-based reduced-order model (ROM) for unsteady aerodynamics and aeroelasticity. To effectively capture temporal dependencies inherent in nonlinear unsteady flow phenomena, an architecture termed Kolmogorov–Arnold Gated Recurrent Network (KAGRN) is introduced. By incorporating a recurrent structure and a gating mechanism, the proposed model effectively captures time-delay effects and enables the selective control and preservation of long-term temporal dependencies. This architecture provides high predictive accuracy, good generalization capability, and fast prediction speed. The performance of the model is evaluated using simulations of the NACA (National Advisory Committee for Aeronautics) 64A010 airfoil undergoing harmonic motion and limit cycle oscillations in transonic flow conditions. Results demonstrate that the proposed model can not only accurately and efficiently predict unsteady aerodynamic coefficients, but also effectively capture nonlinear aeroelastic responses.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Model Reduction and Neural Networks · Fluid Dynamics and Turbulent Flows · Probabilistic and Robust Engineering Design
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Institute of Mechanics, Chinese Academy of Sciences, Beijing 100190, China
School of Future Technology, University of Chinese Academy of Sciences, Beijing 100191, China
School of Engineering Science, University of Chinese Academy of Sciences, Beijing 100191, China
Funding
National Natural Science Foundation of China
References (42)
Show source-provided references
1. Computational fluid dynamics-based transonic flutter suppression with control delay10.1016/j.jfluidstructs.2016.07.002 ↗
2. Design-oriented computational fluid dynamics-based unsteady aerodynamics for flight-vehicle aeroelastic shape optimization10.2514/1.J054024 ↗
3. Two efficient staggered algorithms for the serial and parallel solution of three-dimensional nonlinear transient aeroelastic problems10.1016/S0045-7825(99)00206-6 ↗
4. Transonic limit cycle flutter of high-aspect-ratio swept wings10.2514/1.29547 ↗
5. Robustness and Efficiency of Encapsulated Selective Frequency Damping Using Different Operator-Splitting Schemes: Application to Laminar Cylinder Flow and Transonic Buffet10.1016/j.jcp.2025.113968 ↗
6. Dean, J., Morton, S., McDaniel, D., Clifton, J., and Bodkin, D. (2008, January 18–21). Aircraft stability and control characteristics determined by system identification of CFD simulations. Proceedings of the AIAA Atmospheric Flight Mechanics Conference and Exhibit, Honolulu, HI, USA.10.2514/6.2008-6378 ↗
7. Reduced order unsteady aerodynamic modeling for stability and control analysis using computational fluid dynamics10.1016/j.paerosci.2014.09.001 ↗
8. Reduced-order-model-based flutter analysis at high angle of attack10.2514/1.32285 ↗
9. Accelerating computational fluid dynamics based aeroelastic predictions using system identification10.2514/2.2737 ↗
10. Roy, A., and Mukherjee, R. (January, January 11–15,). Time Series Behaviour of Laminar Separation Bubbles at Low Reynolds Number. Proceedings of the AIAA Scitech 2021 Forum, Virtual.10.2514/6.2021-1197 ↗
11. Nonlinear reduced-order modeling for multiple-input/multiple-output aerodynamic systems10.2514/1.J052323 ↗
12. Volterra-series-based nonlinear system modeling and its engineering applications: A state-of-the-art review10.1016/j.ymssp.2016.10.029 ↗
13. Oscillatory shock motion caused by transonic shock boundary-layer interaction10.2514/3.25144 ↗
14. Self-sustained shock oscillations on airfoils at transonic speeds10.1016/S0376-0421(01)00003-3 ↗
15. Numerical investigation of unsteady wave phenomena for transonic airfoil flow10.1016/j.ast.2012.01.009 ↗
16. Machine learning for fluid mechanics10.1146/annurev-fluid-010719-060214 ↗
17. A novel framework for predicting active flow control by combining deep reinforcement learning and masked deep neural network10.1063/5.0194264 ↗
18. Data-driven modeling for unsteady aerodynamics and aeroelasticity10.1016/j.paerosci.2021.100725 ↗
19. Efficient method for limit cycle flutter analysis based on nonlinear aerodynamic reduced-order models10.2514/1.J050581 ↗
20. Nonlinear aerodynamic reduced-order model for limit-cycle oscillation and flutter10.2514/1.J054951 ↗
21. Neurofuzzy-model-based unsteady aerodynamic computations across varying freestream conditions10.2514/1.J054892 ↗
22. Nonlinear identification via connected neural networks for unsteady aerodynamic analysis10.1016/j.ast.2018.03.034 ↗
23. Linear and nonlinear combined aerodynamic reduced order model based on residual network framework10.1209/0295-5075/ac765e ↗
24. Deep neural network for unsteady aerodynamic and aeroelastic modeling across multiple Mach numbers10.1007/s11071-019-04915-9 ↗
25. Stall flutter prediction based on multi-layer GRU neural network10.1016/j.cja.2022.07.011 ↗
26. Liu, Z., Wang, Y., Vaidya, S., Ruehle, F., Halverson, J., Soljačić, M., Hou, T., and Tegmark, M. (2024). Kan: Kolmogorov-arnold networks. arXiv.
27. Liu, Z., Ma, P., Wang, Y., Matusik, W., and Tegmark, M. (2024). Kan 2.0: Kolmogorov-arnold networks meet science. arXiv.
28. Ji, T., Hou, Y., and Zhang, D. (2024). A comprehensive survey on Kolmogorov Arnold networks (KAN). arXiv.
29. Polo-Molina, A., Alfaya, D., and Portela, J. (2024). MonoKAN: Certified Monotonic Kolmogorov-Arnold Network. arXiv.
30. KAN-ODEs: Kolmogorov–Arnold network ordinary differential equations for learning dynamical systems and hidden physics10.1016/j.cma.2024.117397 ↗
31. Aghaei, A.A. (2024). KANtrol: A Physics-Informed Kolmogorov-Arnold Network Framework for Solving Multi-Dimensional and Fractional Optimal Control Problems. arXiv.
32. Genet, R., and Inzirillo, H. (2024). Tkan: Temporal kolmogorov-arnold networks. arXiv.10.2139/ssrn.4825654 ↗
33. Xu, K., Chen, L., and Wang, S. (2024). Kolmogorov-arnold networks for time series: Bridging predictive power and interpretability. arXiv.
34. Zhou, Q., Pei, C., Sun, F., Han, J., Gao, Z., Pei, D., Zhang, H., Xie, G., and Li, J. (2024). KAN-AD: Time series anomaly detection with Kolmogorov-Arnold networks. arXiv.
35. Tan, Y. (2023). Numerical Analysis and Suppression of Transonic Airfoil Flutter. [Master’s Thesis, University of Chinese Academy of Sciences].
36. NACA 0012 oscillatory and transient pitching
37. On the transonic-dip mechanism of flutter of a sweptback wing10.2514/3.61226 ↗
38. Time series data prediction using sliding window based RBF neural network
39. Deep learning based reduced order model for airfoil-gust and aeroelastic interaction10.2514/1.J059027 ↗
40. An evaluation of data-driven identification strategies for complex nonlinear dynamic systems10.1007/s11071-016-2761-x ↗
41. Nonlinear inviscid aerodynamic effects on transonic divergence, flutter, and limit-cycle oscillations10.2514/2.1720 ↗
42. Rivera, J., Dansberry, B., Bennett, R., Durham, M., and Silva, A.W. (1992, January 13–15). NACA 0012 benchmark model experimental flutter results with unsteady pressure distributions. Proceedings of the 33rd Structures, Structural Dynamics and Materials Conference, Dallas, TX, USA.10.2514/6.1992-2396 ↗